0 6614 2 PDF

Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

FHWA/TX-13/0-6614-2
4. Title and Subtitle 5. Report Date
CHARACTERIZATION AND BEST USE OF RECYCLED ASPHALT October 2012
SHINGLES IN HOT-MIX ASPHALT Published: July 2013
6. Performing Organization Code
7. Author(s) 8. Performing Organization Report No.

Fujie Zhou, Hongsheng Li, Sheng Hu, Joe W. Button, and Jon A. Epps Report 0-6614-2
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Texas A&M Transportation Institute
College Station, Texas 77843-3135 11. Contract or Grant No.
Project 0-6614
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Texas Department of Transportation Technical Report:
Research and Technology Implementation Office September 2011–August 2012
P. O. Box 5080
14. Sponsoring Agency Code
Austin, Texas 78763-5080
15. Supplementary Notes
Project performed in cooperation with the Texas Department of Transportation and the Federal Highway Administration.
Project Title: Use of Recycled Asphalt Shingles in HMA
URL: http://tti.tamu.edu/documents/0-6614-2.pdf
16. Abstract
Recycled asphalt shingles (RAS) often containing more than 20 percent asphalt binder has become another black gold
in the asphalt industry. There are two basic types of RAS scraps in the market: tear-off asphalt shingles (TOAS) and
manufacture waste asphalt shingles (MWAS). Both TOAS and MWAS have been used in hot-mix asphalt (HMA) paving in
Texas. Since RAS binder is very stiff, lots of concerns have been raised on using RAS in HMA. The researchers conducted a
comprehensive investigation on HMA mixes containing RAS, including RAS binder characterization and blending charts for
virgin/RAS binders, impact of RAS content on optimum asphalt content (OAC) and engineering properties of RAS mixes, and
approaches for improving cracking resistance of RAS mixes. Furthermore, a variety of RAS field test sections were constructed
to validate the approaches for improving cracking resistance of RAS mixes. Additionally, this report discusses the
environmental and economic benefit of using RAS in HMA. Based on the research performed, it was found that TOAS binders
with an average of high temperature grade of 175°C are much stiffer than MWAS binders, which have an average of high
temperature grade of 131°C. Impact of MWAS binder on the high and low PG temperatures of virgin binder is different from
that of TOAS binders. Thus it is important to consider differentiating MWAS from TOAS when used in asphalt mixes. The
use of RAS has no significant influence on dynamic moduli of HMA mixes, but improves their rutting/moisture damage.
However, RAS mixes have very poor cracking resistance, compared with the 0 percent RAS mixes with PG64-22 or PG70-22,
even though the RAS mixes have higher OAC. Two approaches for improving cracking resistance of RAS mixes were
explored in the laboratory. The test results clearly indicated that both using soft binder and increasing design density can
improve cracking resistance of RAS mixes. When considering rutting/moisture damage of RAS mixes, using soft binder is
superior to increasing design density. Furthermore, a series of field test sections have been constructed around Texas. The
observed field performance of the test sections indicated that RAS mixes can be successfully used as surface layer of new
construction with strong foundation. It also has been observed that the use of extra virgin asphalt binder (or increasing design
density) can effectively improve cracking performance of RAS mixes. Apparently, more work is also needed to investigate the
performance of RAS/RAP mixes produced at the warm mix temperatures.
17. Key Words 18. Distribution Statement
Recycled Asphalt Shingle, Hot-Mix Asphalt, Mix No restrictions. This document is available to the public through
Design NTIS:
National Technical Information Service
Alexandria, Virginia 22312
http://www.ntis.gov
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 88
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
CHARACTERIZATION AND BEST USE OF
RECYCLED ASPHALT SHINGLES IN HOT-MIX ASPHALT
by
Fujie Zhou, Ph.D., P.E.

Associate Research Engineer
Hongsheng Li
Research Associate
Sheng Hu, Ph.D., P.E.

Assistant Research Engineer
Joe W. Button, P.E.

Senior Research Fellow
and
Jon A. Epps, P.E.

Executive Associate Director
Report 0-6614-2
Project 0-6614
Project Title: Use of Recycled Asphalt Shingles in HMA
Performed in cooperation with the

Texas Department of Transportation
and the
Federal Highway Administration
October 2012
Published: July 2013
TEXAS A&M TRANSPORTATION INSTITUTE

College Station, Texas 77843-3135
DISCLAIMER
The contents of this report reflect the views of the authors, who are responsible for the facts
and the accuracy of the data presented here. The contents do not necessarily reflect the official
view or policies of the Texas Department of Transportation (TxDOT) or the Federal Highway
Administration (FHWA). This report does not constitute a standard, specification, or regulation.
The engineer in charge was Dr. Fujie Zhou, P.E. (Texas, # 95969).
There is no invention or discovery conceived or first actually reduced to practice in the
course of or under this contract, including any art, method, process, machine, manufacture, design
or composition of matter, or any new useful improvement thereof, or any variety of plant, which is
or may be patentable under the patent laws of the United States of America or any foreign country.
The United States Government and the State of Texas do not endorse products or
manufacturers. Trade or manufacturers’ names appear herein solely because they are considered
essential to the object of this report.
v
ACKNOWLEDGMENTS
This project was made possible by the Texas Department of Transportation in cooperation
with the Federal Highway Administration. The authors thank the many personnel who contributed
to the coordination and accomplishment of the work presented here. Special thanks are extended
to Robert Lee, P.E., for serving as the project director. Many people volunteered their time to
serve as project advisors, including:
• Jerry Peterson.
• Steve Smith.
• Gisel Carrasco.
• Woody Raine.
• Mike Wittie.
• Mary Fletcher.
vi
TABLE OF CONTENTS
List of Figures ................................................................................................................................. viii
List of Tables ...................................................................................................................................... x
Chapter 1 Introduction ........................................................................................................................ 1
Chapter 2 RAS Binder Characterization: MWAS vs. TOAS ............................................................. 3
Validation of Asphalt Binder Extraction and Recovery Methods .............................................. 3
RAS Binder Characterization ..................................................................................................... 7
Discussion on Production Temperatures of RAS Mixes .......................................................... 10
Summary and Findings ............................................................................................................. 12
Chapter 3 Development of Virgin and RAS Binders Blending Charts ............................................ 15
Investigation of Virgin and RAS Binders Blending ................................................................. 15
Evaluation of Blending among Virgin, RAP and RAS Binders ............................................... 19
Discussion on Maximum Allowable Asphalt Binder Replacement ......................................... 24
Summary and Findings ............................................................................................................. 27
Chapter 4 Laboratory Evaluation of RAS Mix Engineering Properties ........................................... 29
Preliminary Investigation of Lab Compaction Temperature for RAS Mixes .......................... 29
Impact of RAS on Mix Properties ............................................................................................ 31
Investigation of Actual Blending between Virgin and RAS Binder through Measuring
Engineering Properties ................................................................................................. 36
Summary and Discussion ......................................................................................................... 43
Chapter 5 Approaches for Improving Cracking Resistance of RAS Mixes ..................................... 45
Use of Soft Binders................................................................................................................... 45
Increasing Design Density ........................................................................................................ 47
Summary................................................................................................................................... 48
Chapter 6 Performance of Field Test Sections with RAS Mixes ..................................................... 51
RAS Test Sections on US87, Amarillo and Observed Field Performance ............................... 51
Field Test Section on SH146, Houston and Observed Performance ........................................ 52
Field Test Sections on FM973, Austin, Texas, and Associated Performance .......................... 53
RAP/RAS Field Test Sections on Loop 820, Fort Worth......................................................... 55
Summary................................................................................................................................... 56
Chapter 7 Environmental and Cost Impacts of RAS on Asphalt Mixtures ...................................... 59
Introduction .............................................................................................................................. 59
Environmental Issues ................................................................................................................ 60
Economic Factors ..................................................................................................................... 64
Pertinent Activities to Encourage Recycling ............................................................................ 70
Summary and Conclusions ....................................................................................................... 71
Chapter 8 Summary and Conclusions............................................................................................... 73
References ......................................................................................................................................... 75
vii
LIST OF FIGURES
Figure 1. Flow Chart of Validation of Binder Extraction and Recovery Methods............................ 4

Figure 2. FTIR Test: Original Shingles Binder before and after Extraction and Recovery. ............. 5
Figure 3. FTIR Test: Trichloroethylene. ............................................................................................ 5
Figure 4. Rheological Properties of Shingles Binder: before and after Extraction/Recovery. .......... 6
Figure 5. BBR Test: Early Fractured RAS Binder Beam. ................................................................. 8
Figure 6. BBR Test: Overly Deformed RAS Binder Beam............................................................... 9
Figure 7. High Temperature Grades of RAS Binders: MWAS and TOAS. .................................... 10
Figure 8. RAS Binder Transfer to Limestone Aggregates at Four Different Temperatures. .......... 12
Figure 9. Illustration of Coating during Production: Virgin vs. RAS/RAP Mixes.......................... 12
Figure 10. Binder Blending between PG64-22-A and TOAS-E Binder.......................................... 16
Figure 11. Binder Blending between PG64-28 and TOAS-A Binder. ............................................ 17
Figure 12. Binder Blending between PG64-22-B and MWAS-C Binder........................................ 18
Figure 13. Binder Blending between PG64-22-B and MWAS-A Binder. ...................................... 19
Figure 14. Binder Blending with Fixing 20 Percent TOAS-E Binder and Varying PG64-22-
A and RAP-A Binder. .......................................................................................................... 21
Figure 15. Binder Blending with Fixing 20 Percent RAP-A Binder and Varying PG64-22-
A and TOAS-E Binder. ........................................................................................................ 22
Figure 16. Binder Blending with Fixing 5 Percent MWAS-A Binder and Varying PG64-22-
B and RAP-B Binder. .......................................................................................................... 23
Figure 17. Binder Blending with Fixing 10 Percent RAP-B Binder and Varying PG64-22-B
and MWAS-A Binder. ......................................................................................................... 24
Figure 18. Illustration of Linear Blending Charts for Virgin/RAP/RAS Binders. .......................... 26
Figure 19. Hamburg and OT Testing Results: Field Core vs. Lab-Mixed and Compacted
Specimens at 3 Compaction Temperatures. ......................................................................... 30
Figure 20. Gradations of HMA Mixes with Varying RAS Contents............................................... 32
Figure 21. Impact of RAS on Dynamic Moduli of HMA Mixes..................................................... 34
Figure 22. Impact of RAS on Rutting/Moisture Damage of HMA Mixes. ..................................... 35
Figure 23. Impact of RAS on Cracking Resistance of HMA Mixes. .............................................. 36
Figure 24. Retained Binder Contents for the RAP and Virgin Aggregates after Blending. ............ 37
Figure 25. Measured E* Values of 28 Percent RAP Specimens Produced under Different
Conditions. ........................................................................................................................... 39
Figure 26. Measured E* Values of 5 Percent RAS Specimens Produced under Different
Conditions. ........................................................................................................................... 40
Figure 27. Master Curve of Extracted/Recovered Binder. .............................................................. 40
Figure 28. Measured vs. Predicted E* Values. ................................................................................ 41
Figure 29. Illustration of Different Mixing and Molding Processes................................................ 42
Figure 30. Field Core vs. Lab Normal Blending vs. Lab 100 Percent Blending. ............................ 42
Figure 31. Impact of Soft Binders on Dynamic Modulus of 5 Percent RAS Mixes. ...................... 46
Figure 32. Impact of Soft Binders on Rutting/Moisture Damage of 5 Percent RAS Mixes. .......... 47
Figure 33. Impact of Soft Binder on Cracking Resistance of 5 Percent RAS Mixes. ..................... 47
Figure 34. Impact of Increasing Design Density on Rutting/Moisture Damage and Cracking
Resistance of RAS Mixes. ................................................................................................... 48
Figure 35. Observed Reflective Cracking of RAS Test Sections on US87, Amarillo, Texas. ........ 52
viii
Figure 36. Reflective Cracking Development History of RAS Test Sections on US87,
Amarillo, Texas. .................................................................................................................. 52
Figure 37. Perfect Condition of RAP/RAS Test Section on SH146, Houston. ............................... 53
Figure 38. Pavement Conditions of RAP/RAS Test Sections 3 and 6 on FM973, Austin,
Texas, in July 2012. ............................................................................................................. 55
Figure 39. Existing CRCP Condition and Construction of RAP/RAS Test Sections on Loop
820, Fort Worth.................................................................................................................... 56
Figure 40. Price Trends for Liquid Asphalt from 1999 to 2011 (after Peterson, 2011). ................. 64
Figure 41. HMA Cost in Dollars per Ton for Type D (1/2-inch NMAS) Mix Containing PG
76-22. ................................................................................................................................... 70
ix
LIST OF TABLES
Table 1. Six RAS Mixes Investigated under This Study. ................................................................ 31

Table 2. OAC of Each RAS Mix. .................................................................................................... 33
Table 3. RAS Mixes with Soft Virgin Binders. ............................................................................... 45
Table 4. Nine Test Sections on FM973, Austin, Texas. .................................................................. 54
Table 5. Four Field Test Section on Loop 820. ............................................................................... 55
Table 6. Asbestos-Containing Asphalt Roofing Products (after Townsend et al., 2007). ............... 61
Table 7. Typical Shingle Composition (modified after Brock, 2007). ............................................ 65
Table 8. Typical HMA Plant Economic Savings When Using RAS (after Brock, 2007). .............. 66
Table 9. Method for Calculating the Value of RAS in Asphalt Mixtures (modified after
Hughes, 1997; Hansen, 2009; and Krivit, 2007). ................................................................ 68
Table 10. Assumptions Used for Asphalt Pavement Cost Estimates. ............................................. 69
Table 11. Asphalt Pavement Cost Estimates. .................................................................................. 69
x
CHAPTER 1
INTRODUCTION
Reclaimed asphalt pavement (RAP) has been one of the most often used recycling
materials in the asphalt industry. With increases in the price of asphalt cement and subsequent
price fluctuations, the industry has further amplified its recycling efforts. Most recently, the use
of recycled asphalt shingles (RAS) in hot-mix asphalt (HMA) has become another black gold to
the asphalt paving industry since RAS contains usually 20–30 percent asphalt binder of its total
weight. In addition to conserving energy and protecting the environment, the use of RAS can
significantly reduce the increasing cost of HMA paving.
There are two basic types of RAS scraps in the market: tear-off asphalt shingles (TOAS)
and manufacture waste asphalt shingles (MWAS). In United States, around 10 million tons of
TOAS and 1 million tons of MWAS are available for recycling. Specifically, there are several
national wide, large roof shingles manufactories in Texas, such as Owens Corning, GAF,
TAMKO, Certain Teed, etc. Significant markets exist for both recycling and paving industries.
More than 30 years ago, some of the original pioneers established the first shingles
recycling plants, investigated mix designs containing RAS, and then published the first technical
literature in the late 1980s (Epps and Paulsen, 1986; Paulsen et al., 1986; Shepherd et al., 1989).
More recently, additional HMA producers, departments of transportation (DOTs), and
researchers further expanded the expertise in shingle recycling in HMA (Grzybowski, 1993;
Newcomb et al., 1993; Button et al., 1996; Janisch and Turgeon, 1996; Foo et al., 1999; NAHB,
1998; Dykes, 2002; Lum, 2006; Brock 2007; McGraw et al., 2007; Schroer, 2007, Johnson et al.,
2010; Williams et al. 2011). All these efforts paved the way for more DOTs to use RAS in
HMA.
In February 2009, the Texas Commission on Environmental Quality (TCEQ) issued an
Authorization Memo to allow HMA plants to include either MWAS or TOAS under the TCEQ
air quality standard permit for permanent HMA plants. Since then, RAS has been widely used in
Texas. Meanwhile, some concerns on consistency of processed RAS, stiffness of RAS binder,
and durability of RAS mixes were raised. To address these concerns, in 2010 the Texas
Department of Transportation (TxDOT) initiated a research study (Project 0-6614) at the
Texas A&M Transportation Institute (TTI) with an overall objective of improving the use of
RAS in HMA. To achieve this main objective, the following steps were undertaken by the
researchers:
• Develop best practices for use of RAS in HMA.
• Characterize RAS binder properties.
• Develop virgin/RAS binders blending charts.
• Identify the impact of RAS on mix engineering properties.
• Evaluate approaches for improving durability of RAS mixes in the lab.
• Construct field test sections to validate the approaches for improving durability of RAS
mixes.
Except the best practices for using of RAS in HMA documented in Report 0-6614-1,
details of all others are presented in Chapters 2, 3, 4, 5, and 6 of this report, respectively.
1
Additionally, the environmental and cost impacts of RAS on asphalt mixes are discussed in
Chapter 7. This is followed by summary and conclusions at the end of the report.
2
CHAPTER 2
RAS BINDER CHARACTERIZATION: MWAS VS. TOAS
The coating asphalt used for making new roof shingles is very stiff, but it is seldom
known that such stiff asphalt is made from an asphalt flux (i.e., asphalt cement [AC]-5) by a
process known as air blowing process. During the process, air is bubbled through a large tank
containing the hot flux. Heat and oxygen cause chemical reaction that changes the
characteristics of the asphalt. The whole process is monitored, and the blowing is stopped when
the correct properties are reached. Thus the coating asphalt is highly oxidized in the beginning,
and it becomes even stiffer and more oxidized for sitting on the roof for 20-30 years under the
sun. Up to now the information on RAS binders, in terms of performance grade (PG), has been
very limited. One of the reasons is that the roofing industry is still using Penetration and Ring
and Ball soft point to grade the coating asphalt. Another reason is due to limitation of existing
dynamic shear rheometer (DSR). This chapter will grade the binders extracted and recovered
binders from MWAS and TOAS following the Superpave PG system.
VALIDATION OF ASPHALT BINDER EXTRACTION AND RECOVERY METHODS
Solvent-based asphalt binder extraction and recovery become necessary if one needs to
characterize the recovered binders. However, there are always concerns especially about the
solvent-based asphalt binder recovery process. One of the concerns is that the properties of the
recovered asphalt binder may be changed for two potential reasons: 1) some solvent left in the
recovered asphalt binder (note that solvent often softens asphalt binder), or 2) the recovered
asphalt binder is stiffened due to over-cooking for removing the solvent. To address this concern
and validate the extraction and recovery methods used, TTI researchers compared both
rheological properties and chemical components of one original shingles binder with the
extracted/recovered binder from the MWAS produced with the same original shingles binder
tested. The rheological properties were evaluated using DSR and bending beam rheometer
(BBR), and the chemical property was measured with Fourier transform infrared spectroscopy
(FTIR). Note that the original shingles binder was directly received from a binder supplier and
no filler was added. Figure 1 shows the whole process. The solvent used in binder extraction is
trichloroethylene. The extraction and recovery methods employed in this study are:
• Tex-210-F Determining Asphalt Content of Bituminous Mixtures by Extraction: Part I-
Centrifuge Extraction Method Using Chlorinated Solvent.
• ASTM D5404 Standard Practice for Recovery of Asphalt from Solution Using the Rotary
Evaporator.
3
Original shingles binder
MWAS produced with the Solvent:

same original shingle binder Trichloroethylene
Binder extraction and recovery:

Tex-210-F and ASTM D5404
DSR/BBR Extracted/recovered binders FTIR
Before=after
No No Before=after
Need more research extraction/recovery
extraction/recovery
Yes Yes
Tex-210-F and ASTM D5404
Figure 1. Flow Chart of Validation of Binder Extraction and Recovery Methods.
Chemical Property Comparison of RAS Binder before and after Extraction and Recovery
Figure 2 presents the FTIR test results of original shingles binder before and after the
extraction and recovery. The chemical components of the original shingles binder are exactly
the same as those of the extracted and recovered shingles binder. Furthermore, Figure 3 also
shows the FTIR test result of the trichloroethylene solvent itself. Apparently, the
trichloroethylene itself has large absorbance when wavelength is less than 1000 (cm-1). If there
is any trichloroethylene left, the absorbance values of the recovered shingles binder will be
different from those of original shingles binder in that wavelength range. Therefore, the
trichloroethylene solvent was completely removed during the recovery process.
4
FTIR: Original Shingles Binder-before and after Extraction
and Recovery
0.40
Before Extraction and Recovery
0.35
After Extraction and Recovery
0.30
0.25
Absorbance
0.20
0.15
0.10
0.05
0.00
500 1000 1500 2000 2500 3000 3500 4000
Wavelength (cm-1)
Figure 2. FTIR Test: Original Shingles Binder before and after Extraction and Recovery.
FTIR: Trichloroethylene
0.9
0.8
0.7
0.6
Absorbance
0.5
0.4
0.3
0.2
0.1
0
500 1000 1500 2000 2500 3000 3500 4000
Wavelength (cm-1)
Figure 3. FTIR Test: Trichloroethylene.
Rheological Property Comparison of RAS Binder before and after Extraction and
Recovery
Figure 4 illustrates the DSR and BBR test results of the original shingles binder before
and after the extraction and recovery. Note that the BBR beams were prepared using the
pressure aging vessel (PAV) aged asphalt binder residue. The rheological properties of the
shingles binder, in terms of PG grade, have no change before and after the extraction and
recovery. Note that the low temperatures grade of the PAV aged shingles binder are beyond the
limits of BBR test so no data are available (more discussion in later section). Instead, the S and
5
m values at 0°C were used for comparison. Again, the extraction and recovery process did not
change the rheological properties of the shingles binder. With this validation, the authors
conducted extensive shingles binder extraction and recovery following the Tex-210-F and
ASTM D5404, and then evaluated the RAS binder properties, as discussed in next section.
DSR: Original Shingles Binder

140
135
High Temp. Grade (°C)
130
125
120
115
110
105
100
Before Extraction/Recovery After Extraction/Recovery
BBR@0°C: PAV Aged Shingles Binder

45
Stiffness (MPa)@60second
40
35
30
25
20
15
10
5
0
BBR@0°C: PAV Aged Shingles Binder

0.25
0.20
m value@60second
0.15
0.10
0.05
0.00
Figure 4. Rheological Properties of Shingles Binder: before and after Extraction/Recovery.
6
RAS BINDER CHARACTERIZATION
As discussed previously, roofing shingle binders, regardless of MWAS or TOAS, are

very stiff, and they are far stiffer than any PG76-22 binder. Note that PG76-22 binders are the
stiffness virgin binder being used in Texas. It is important to know the true grade of the RAS
binder since it has significant influence on virgin binder selection and then the allowable,
maximum amount of RAS used in the asphalt mixes. This study sampled, extracted, recovered,
and characterized a variety of RAS binders. Detailed information is presented below.
Selection of RAS Samples
A variety of processed RAS including both MWAS and TOAS were collected from
contractors and shingles recyclers. In Texas major shingles manufacturers are mainly located in
Dallas-Fort Worth and Houston areas. The contractors in these areas have been using MWAS
collected from those shingles manufactories in the last three years. Four different types of
MWAS often used in Texas were sampled from the processed MWAS stockpiles in this study,
and are designated with MWAS-A, MWAS-B, MWAS-C, and MWAS-D. Additionally, six
TOAS were also selected from different contractors and recyclers around Texas with the
following designation: TOAS-A, TOAS-B, TOAS-C, TOAS-D, TOAS-E, and TOAS-F. In
summary, a total of 10 processed RAS were selected and evaluated under this study.
RAS Binder Extraction and Recovery
RAS binders were extracted and recovered from the 10 processed RAS selected above,
following the validated extraction and recovery methods (Tex-210-F Part I and ASTM D5404).
There was some difficulty in draining out of the recovered TOAS binders, which were so stiff
that they just did not flow out of the beaker even at 165°C (329°F) after finishing the recovery
process. In one case, the oven temperature was raised up to 200°C (392°F) in order to drain out
the TOAS binder.
RAS Binder Characterization
Both DSR and BBR were used to grade the 10 extracted/recovered RAS binders. The
results are discussed as follows.
• BBR test results
As noted previously, the researchers had difficulty in grading PAV aged shingles binder
using BBR. There are two criteria (S and m) for determining the low temperature grade
of asphalt binders. It is no problem for RAS binders to meet the S (<300 MPa) criteria,
but the measured m values were always less than 0.3. Some RAS binder beams fractured
even before reaching 240 seconds (see Figure 5). The reason for having such a small m
value is that RAS binders including MWAS binders have much less capability to relax
under strain. Note that the original shingles binders are already substantially oxidized
through air blowing process. The researchers even tried to run the BBR test at higher
temperatures (i.e., 18°C and even 24°C), but the measured m values are still less than 0.3,
and in some cases the beam deformation reaches the limit of BBR machine within a very
short of period of time (see Figure 6). Therefore, no reliable results from BBR test were
7
obtained for any one of the 10 recovered RAS binders. Alternative tests (such as Asphalt
Binder Cracking Device test) should be explored.
Figure 5. BBR Test: Early Fractured RAS Binder Beam.
8
Figure 6. BBR Test: Overly Deformed RAS Binder Beam.
• DSR test results

To measure the high temperature grades of those extremely stiff binders, a high
temperature DSR was specifically purchased under this study. Nine of the 10 RAS
binders were successfully graded following Superpave binder PG system. The high
temperature grade of one TOAS binder is beyond the upper limit of the purchased DSR,
which is 200°C, so that extrapolation was used to estimate its high temperature grade.
For each extracted/recovered RAS binder, both original and rolling thin-film oven aged
residue were evaluated. The high temperature grades of the 10 RAS binders are shown in
Figure 7.
Several observations can be clearly made from Figure 7:
o TOAS binders with an average of high temperature grade of 175°C (347°F) are
much stiffer than MWAS binders, which have an average of high temperature
grade of 131°C (268°F).
o Compared to the TOAS varying from 159°C (318°F) to 214°C (417°F), the
MWAS has smaller variation in terms of the high temperatures grade.
These two observations clearly indicate that the MWAS is different from the TOAS. It is
necessary and important to differentiate the MWAS from the TOAS when used in asphalt mixes.
For example, TxDOT may consider allowing smaller amount of TOAS in the specification when
compared with MWAS.
9
230
Tear-offs
210 Average PG=178
CV=10.8%
Manufacture Waste
190
Average PG=131
High Temp. Grade (°C )
170 CV=5.5%
150
130
110
90
70
50
MWAS-A MWAS-B MWAS-C MWAS-D TOAS-A TOAS-B TOAS-C TOAS-D TOAS-E TOAS-F
Figure 7. High Temperature Grades of RAS Binders: MWAS and TOAS.
DISCUSSION ON PRODUCTION TEMPERATURES OF RAS MIXES
Production temperature of asphalt mixes depends on the viscosity of asphalt binders and
how well the asphalt binder coats the aggregates. For all virgin mixes without RAP/RAS,
general rules have been well established. For example, the production temperature for a virgin
mix with a PG64-22 binder is around 143°C (290°F) at which the virgin binder PG64-22 can
easily flow and then coat virgin aggregates. However, the mixing (or production) temperature for
RAS (or RAP) mixes has not well established. This study explored how mixing (or production)
temperature affects RAS binder melting and coating virgin aggregates. The researchers blended
white limestone aggregates with a MWAS and a TOAS at different temperatures, respectively.
Four mixing (or production) temperatures used in the lab are 143°C (290°F), 149°C (300°F),
163°C (325°F), and 200°C (392°F). The process of the mixing temperature investigation is
described below:
• Step 1: Sieve limestone aggregates and keep the aggregates passing 1/2 inch sieve and
retaining on 3/8 inch sieve.
• Step 2: Wash the aggregates and dry them up in oven.
• Step 3: Weight the dried limestone aggregate and RAS (either MWAS or TOAS) in a
ratio of 80:20.
• Step 4: Load the shallow pan with the limestone aggregates into an oven with pre-set
temperature (i.e., 143°C [290°F]) and store them overnight; Meanwhile, store the RAS in
a hot room with the temperature of 60°C (140°F) overnight.
• Step 5: In the second morning, manually blend the aggregates with RAS and then store
them together for 2 hrs at the same oven with the same pre-set temperature as that for
aggregates. The purpose of this step is to heat up the RAS and make sure that the RAS
and limestone aggregates have the same pre-set temperature (i.e., 143°C [290°F]).
10
• Step 6: Without adding any virgin binder, blend the aggregates and RAS using bucket
mixer for 2–3 minutes. Then put the mixed RAS/limestone aggregates back to the same
oven and store for another 2 hrs at the same pre-set temperature.
• Step 7: Separate the RAS from limestone aggregate through sieving with a 3/8 inch sieve.
• Step 8: Take photos to the limestone aggregates and see the change of color of the
limestone.
Figure 8 shows the limestone aggregates after mixing at four different mixing
temperatures. Extremely high mixing (or production) temperature is needed in order to have
MWAS binder (around PG130) to melt and coat virgin aggregates more and better. However,
the TOAS even at the temperature of 200°C (392°F) did not show much melting and coating
effect.
11
Figure 8. RAS Binder Transfer to Limestone Aggregates at Four Different Temperatures.
Referring the mixing temperatures for virgin mixes with virgin binders (i.e., 143°C
[290°F] for PG64-22, 149°C [300°F] for PG70-22, 163°C [325°F] for PG76-22), it seems that
the mixing (or production) temperature needs to be at least two times of its high temperature
grade in order to make the RAS binder flow and coating virgin aggregates. This means that the
production temperature for a mix with a MWAS binder (PG130-XX) should be around 260°C
(500°F). For those mixes with TOAS (PG175-XX) the production temperature may need to go
up to 350°C (662°F). Such high production temperature is impractical. The actually mixing (or
production) temperature for RAP/RAS mixes with PG64-22 binders used in Texas is around
149°C (300°F). Thus, it is unrealistic to expect that the binder in RAS, regardless of MWAS or
TOAS, will melt down and flow out to coat virgin aggregates. The most possible scenario is that
RAS acts like a black rock and has very little (if not none) contribution in the coating process.
The reduced virgin binder (i.e., PG64-22) needs to coat both virgin binder and RAS, as
illustrated in Figure 9. The major difference between virgin mix and RAS (RAP) mix is that the
RAS mix often has thinner film thickness than virgin mix if the RAS binder is assumed to be
100 percent active. The good thing for RAS (or RAP) mix is that the virgin binder can activate
as a rejuvenator to continue to soften the hard RAS binder in the period of pavement service.
RAS coated
with virgin
binder
Virgin Mix RAS Mix
Figure 9. Illustration of Coating during Production: Virgin vs. RAS/RAP Mixes.
SUMMARY AND FINDINGS
Based on the data presented previously, the following summary and findings are offered:
• The asphalt binder extraction and recovery procedures (Tex-210-F Part I and ASTM
D5404) are validated in this study. No solvent was detected through FTIR testing, and
12
the binder rheological properties, in terms of PG high and low temperatures, were almost
the same before and after going through the extraction and recovery processes.
• TOAS binders with an average of high temperature grade of 175°C (347°F) are much
stiffer than MWAS binders, which have an average of high temperature grade of 131°C
(268°F). Compared to the TOAS varying from 159°C (318°F) to 214°C (417°F), the
MWAS has smaller variation in terms of the high temperatures grade. These two
observations clearly indicate that the MWAS is different from the TOAS. It is necessary
and important to differentiate the MWAS from the TOAS when used in asphalt mixes.
In general, the RAS binders are very stiff, and regular DSR and BBR could not
characterize rheological properties of these extremely stiff binders. Therefore it is critical to
investigate the impact of these stiff binders on rheological properties of the combined binder
after blending with virgin binders, which is discussed in next section.
13
CHAPTER 3
DEVELOPMENT OF VIRGIN AND RAS BINDERS BLENDING CHARTS
Many efforts have been made to evaluate the blending between virgin binders and RAP
binders, and all results indicated that the RAP binders linearly blend with virgin binders.
Compared to virgin/RAP binder blending, there was very little work done on virgin/RAS binders
blending in the literature, although the American Association of State Highway and
Transportation Officials (AASHTO) PP53, Standard Practice for Design Consideration when
Using Reclaimed Asphalt Shingle (RAS) in New Hot-Mix Asphalt (HMA), recommends that the
linear blending used for virgin/RAP binders blending also be used with virgin/RAS binders. One
reason, as discussed previously, may be the difficulty in grading RAS binder using regular DSR
and BBR. This study investigated the full blending charts for three virgin binders and four RAS
binders extracted/recovered from both TOAS and MWAS. Additionally, virgin/RAS/RAP
binder blending was examined as well. Detailed information is presented below.
INVESTIGATION OF VIRGIN AND RAS BINDERS BLENDING
Virgin and RAS Binders
Three virgin binders selected for blending are 1) PG64-22-A, 2) PG64-22-B, and 3)
PG64-28, and the four RAS binders are TOAS-A, TOAS-E, MWAS-A, and MWAS-C. With
these selected binders, a total of four combinations of virgin/RAS binders, as listed below, were
evaluated under this study. Note that these four combinations have been used in the field test
sections.
• Virgin Binder: PG64-22-A and RAS Binder: TOAS-E.
• Virgin Binder: PG64-28 and RAS Binder: TOAS-A.
• Virgin Binder: PG64-22-B and RAS Binder: MWAS-A.
• Virgin Binder: PG64-22-B and RAS Binder: MWAS-C.
Laboratory Testing, Results, and Analysis
For each combination, different percentages of virgin and RAS binders were blended and
then evaluated through DSR and BBR testing in terms of the high and low PG temperatures. The
test results for these four combinations are presented in Figures 10, 11, 12, and 13, respectively.
The following observations are made from Figures 10 through 13.
• Generally the virgin and RAS binders blending is non-linear.
• For practically application, the linear blending chart can still be used if the RAS binder
percentage is less than 30 percent. Within 30 percent RAS binder, not only is the linear
blending chart applicable, but the regular DSR and BBR can also be used to evaluate the
high and low PG temperatures of the blended binders.
• Increasing RAS binder will improve the high temperature grade of virgin binder and also
warm up its low temperature grade, which is good for rutting resistance but causes
concerns on cracking resistance of the blended binder. Adding 20 percent RAS binder
can make a PGxx-22 binder become a PGxx-16 (or even a PGxx-10 in Figure 9) binder
after blending. Additionally, the necessity of using PGxx-28 virgin binder can also be
15
seen in order to get a PGxx-22 combined binder when 20 percent RAS binder is added
(Figure 11). Note that 20 percent RAS binder is corresponding to 5 percent RAS in
weight of the total mix with an assumption that the optimum asphalt content of a RAS
mix is 5 percent and RAS contains 20 percent asphalt binder in it.
• Impact of MWAS binders on the high and low PG temperatures of virgin binders is
different from that of TOAS binders. Compared to the TOAS binders (Figures 10 and
11), the MWAS binders (Figures 12 and 13) have less impact on PG temperatures of
virgin binders, which makes sense since TOAS binders are much stiffer than those
MWAS binders (see Figure 7). Therefore, it is necessary to consider differentiating the
MWAS from the TOAS when designing HMA containing RAS.
Binder Blending: PG64-22-A/TOAS-E Binder

180
160
140
120
100
80
60
40
20
0
0 10 20 30 40 50 60 70 80 90 100
RAS Binder (%)
Binder Blending: PG64-22-A/TOAS-E Binder

0
-5
Low Temp. Grade (°C)
-10
-15
-20
-25
0 10 20 30 40 50 60 70 80 90 100
RAS Binder (%)
Figure 10. Binder Blending between PG64-22-A and TOAS-E Binder.
16
Binder Blending: PG64-28 and TOAS-A Binder
180
160
High Temp. Grade (°C) 140
120
100
80
60
40
20
0
0 10 20 30 40 50 60 70 80 90 100
RAS Binder (%)
Binder Blending: PG64-28 and TOAS-A Binder

0
-5
-10
-15
-20
-25
-30
-35
0 10 20 30 40 50 60 70 80 90 100
RAS Binder (%)
Figure 11. Binder Blending between PG64-28 and TOAS-A Binder.
17
Binder Blending: PG64-22-B/MWAS-C Binder
140
120

80
60
40
20
0
0 20 40 60 80 100
RAS Binder (%)
Binder Blending: PG64-22-B/MWAS-C Binder

0
-5
-10
-15
-20
-25
-30
0 20 40 60 80 100
RAS Binder (%)
Figure 12. Binder Blending between PG64-22-B and MWAS-C Binder.
18
Binder Blending: PG64-22-B and MWAS-A Binder
160
140
100
80
60
40
20
0
0 20 40 60 80 100
RAS Binder (%)
Binder Blending: PG64-22-B and MWAS-A Binder

0
-5
-10
-15
-20
-25
0 20 40 60 80 100
RAS Binder (%)
Figure 13. Binder Blending between PG64-22-B and MWAS-A Binder.
EVALUATION OF BLENDING AMONG VIRGIN, RAP AND RAS BINDERS
The use of both RAP and RAS in HMA has become a regular practice in asphalt industry,
so this study also briefly explored the blending among virgin/RAP/RAS binders. The same two
virgin binders (PG64-22-A and PG64-22-B), two RAS binders (MWAS-A and TOAS-E), and
two RAP binders (RAP-A and RAP-B) were selected. Again, four combinations listed below
were evaluated with different percentages of binder contents through DSR and BBR testing.
• TOAS-E RAS Binder (=20 percent of the total binder): varying PG64-22-A and RAP-A.
• RAP-A Binder (=20 percent of the total binder): varying PG64-22-A and TOAS-E.
• MWAS-A Binder (=5 percent of the total binder): varying PG64-22-B and RAP-B.
• RAP-B Binder (=10 percent of the total binder): varying PG64-22-B and MWAS-A.
19
The DSR and BBR test results of these four combinations are shown in Figures 14, 15,
16, and 17, respectively. From these figures the following observations are made:
• As long as RAS binder content is fixed in the blending process, the virgin/RAP binders
follows linear blending line, as seen in Figures 14 and 16. Both high and low
temperatures of PG of the combined binder increases linearly with adding RAP binder.
When RAP binder content is fixed, the virgin/RAS binders blending, again, is non-linear
(see Figures 13 and 15).
• When RAS binder is already blended with virgin binder, adding more RAP binder makes
the blended binder even stiffer. For example, as shown in Figure 14, 20 percent RAS
binder itself already modified the PG64-22-A binder to a PG81-15 binder. Adding any
RAP binder (even 5 percent RAP binder) will worsen the cracking resistance of the
combined binder. The similar finding for fixing RAP binder but adding more RAS
binder to the virgin binder can be observed in Figures 15, 16, and 17.
20
Binder Blending with Fixing 20% TOAS-E Binder:
PG64-22-A and RAP-A Binder
110
80%RAP +20%RAS
105
100
20%PG64-22+60%RAP+20%RAS
95
35%PG64-22+45%RAP +20%RAS
90
60%PG64-22+20%RAP+20%RAS
85
80
80%PG64-22+20%RAS
75
70
65
60
0 10 20 30 40 50 60 70 80 90 100
RAP Binder (%)
Binder Blending with Fixing 20% TOAS-E Binder:

PG64-22-A and RAP-A Binder
0
-5
35%PG64-22+45%RAP +20%RAS
-10
60%PG64-22+20%RAP+20%RAS
50%PG64-22+30%RAP+20%RAS
-15
80%PG64-22+20%RAS
-20
-25
0 5 10 15 20 25 30 35 40 45 50
RAP Binder (%)
Figure 14. Binder Blending with Fixing 20 Percent TOAS-E Binder and Varying PG64-22-
A and RAP-A Binder.
21
Binder Blending with Fixing 20% RAP-A Binder:
PG64-22-A and TOAS-E Binder
160 80%RAS +20%RAP
140
20%PG64-22+60%RAS+20%RAP
120
35%PG64-22+45%RAS +20%RAP
100
50%PG64-22+30%RAS+20%RAP
80
80%PG64-22+20%RAP
60
40
20
0
0 10 20 30 40 50 60 70 80 90
RAS Binder (%)
Binder Blending with Fixing 20% RAP-A Binder:

PG64-22-A and TOAS-E Binder
0
-5
60%PG64-22+20%RAS+20%RAP
-10
65%PG64-22+15%RAS +20%RAP
70%PG64-22+10%RAS+20%RAP
-15
75%PG64-22+5%RAS+20%RAP
-20 80%PG64-22+20%RAP
-25
0 5 10 15 20 25
RAS Binder (%)
Figure 15. Binder Blending with Fixing 20 Percent RAP-A Binder and Varying PG64-22-A
and TOAS-E Binder.
22
Binder Blending with Fixing 5% MWAS-A Binder:
PG64-22-B and RAP-B Binder
120
95%RAP +5%RAS
100
35%PG64-22+60%RAP+5%RAS
80
75%PG64-22+20%RAP+5%RAS
95%PG64-22+5%RAS
60
40
20
0
0 10 20 30 40 50 60 70 80 90 100
RAP Binder (%)
Binder Blending with Fixing 5% MWAS-A Binder:

PG64-22-B and RAP-B Binder
0
-5
-10
55%PG64-22+40%RAP +5%RAS
-15
75%PG64-22+20%RAP+5%RAS
-20
85%PG64-22+10%RAP+5%RAS
95%PG64-22+5%RAS
-25
0 5 10 15 20 25 30 35 40 45
RAP Binder (%)
Figure 16. Binder Blending with Fixing 5 Percent MWAS-A Binder and Varying PG64-22-
B and RAP-B Binder.
23
Binder Blending with Fixing 10% RAP-B Binder:
PG64-22-B and MWAS-A RAS Binder
160
10%RAP +90%RAS
140
High Temp. G Grade (°C)
120
100 50%PG64-22+10%RAP +40%RAS 30%PG64-22+10%RAP+60%RAS
80
70%PG64-22+10%RAP+20%RAS
90%PG64-22+10%RAP
60
40
20
0
0 10 20 30 40 50 60 70 80 90 100
RAS Binder (%)
Binder Blending with Fixing 10% RAP-B Binder:

PG64-22-B and MWAS-A RAS Binder
0
30%PG64-22+10%RAP +60%RAS
-5
-10
50%PG64-22+10%RAP +40%RAS
-15
70%PG64-22+10%RAP+20%RAS
-20
80%PG64-22+10%RAP+10%RAS
90%PG64-22+10%RAP
-25
0 10 20 30 40 50 60 70
RAS Binder (%)
Figure 17. Binder Blending with Fixing 10 Percent RAP-B Binder and Varying PG64-22-B
and MWAS-A Binder.
DISCUSSION ON MAXIMUM ALLOWABLE ASPHALT BINDER REPLACEMENT
The DSR and BBR test results presented above clearly indicated that RAS binders have
significant influence on both high and low PG temperatures of the blended binder; positive effect
on high temperature property (or rutting resistance) and negative effect on low temperature
property (or cracking resistance). Thus the maximum allowable RAS binder replacement is
controlled by the influence of RAS binder on the low temperature property of the blended binder.
The virgin/RAS binders blending charts shown in Figures 5 and 6, clearly indicate that the
maximum allowable binder replacement for MWAS should be different from that for TOAS. It
seems OK to use a maximum of 20 percent MWAS binder replacement, but for TOAS binders,
24
the binder replacement should be significantly reduced, depending on the combination of virgin
and TOAS binders.
When both RAP and RAS are used, the maximum allowable recycled binder (RAP binder
plus RAS binder) replacement is influenced by many factors (such as virgin binder, RAP binder,
RAS binder, pavement layers [surface or base layer], climate, traffic, etc.). It will be safe to
directly evaluate the blending chart for virgin, RAP, and RAS binders. Reviewing the impact of
RAS binder on low temperature property of the blended binders shown in Figures 10, 11, 12, 13,
15, and 17, RAS binders, for practical applications, should be limited within 30 percent of the
total binder. If this is the case, the blending chart for virgin/RAP/RAS can be significantly
simplified:
• Linear blending chart is practically applicable to estimate the high and low PG
temperatures of the blended virgin/RAS binders or virgin/RAP/RAS binders, which
significantly reduces the DSR/BBR testing workload, because only the properties of
blended binders at two ends are required, and anything in between can be linearly
interpolated.
For virgin/RAS binders blending (see Figure 18a), one only needs to determine the PG
temperatures of virgin binder and the 30 percent RAS/70 percent virgin binders,
respectively. Then one can determine the continuous PG temperatures of any blending
through linear interpolation. For virgin/RAP/RAS binders blending (see Figures 18b and
18c), similar approach can be used for 1) fixing RAS binder replacement (say 20 percent)
and varying virgin/RAP binders and 2) fixing RAP binder replacement (15 percent) and
varying virgin/RAS binders. Note that RAS/RAP binder replacements shown in
Figure 18b/c are just for demonstration only and can be replaced with real numbers.
• Additionally, the use of the linear blending chart and practical amount of RAP/RAS
binders makes it possible to employ regular DSR and BBR test equipment to evaluate the
properties of the blended virgin/RAP/RAS binders.
25
Blending Chart: Virgin/RAS Binder
-4
30RAS/70Virgin

-10
-16
-22
Virgin Binder
-28
0 5 10 15 20 25 30
RAS Binder (%)
(a)
Blending Chart with Fixing 20%RAS Binder:
Varying Virgin/RAP Binder
-10
-16
20RAS/50RAP/30Virgin
-22
20RAS/80Virgin
-28
-34
0 10 20 30 40 50
RAP Binder (%)
(b)
Blending Chart with Fixing 15%RAP Binder:
Varying Virgin/RAS Binder
-10
-16
-22 15RAP/30RAS/55Virgin
-28 15RAP/85Virgin
-34
0 5 10 15 20 25 30
RAS Binder (%)
(c)
Figure 18. Illustration of Linear Blending Charts for Virgin/RAP/RAS Binders.
26
SUMMARY AND FINDINGS
Based on the data presented previously, the following summary and findings are offered:
• Increasing RAS binder will improve the high temperature grade of virgin binder, and also
warm up its low temperature grade, which is good for rutting resistance but causes
concerns on cracking resistance of the blended binder. Adding 20 percent RAS binder
can make a PGxx-22 binder become a PGxx-16 (or even a PGxx-10 in Figure 10) binder
after blending. Additionally, the necessity of using PGxx-28 virgin binder can also be
seen in order to get a PGxx-22 combined binder when 20 percent RAS binder is added
(Figure 11). Note that 20 percent RAS binder is corresponding to 5 percent RAS in
weight of the total mix with an assumption that the optimum asphalt content of a RAS
mix is 5 percent and RAS contains 20 percent asphalt binder in it.
• Impact of MWAS binders on the high and low PG temperatures of virgin binders is
different from that of TOAS binders. Compared to the TOAS binders (Figures 10 and
11), the MWAS binders (Figures 13 and 14) have less impact on PG temperatures of
virgin binders, which makes sense since TOAS binders are much stiffer than those
MWAS binders (see Figure 7). Therefore, it is necessary to consider differentiating the
MWAS from the TOAS when designing HMA containing RAS.
• Different from virgin/RAP binders blending, the virgin/RAS binder blending, in general,
is non-linear. However, for practical application, the linear blending chart can still be
used for estimating continuous grade (high and low temperatures) of both virgin/RAS
blended binders and virgin/RAP/RAS blended binders, if the RAS binder is limited
within 30 percent of the total binder. In such way, the DSR/BBR testing is significantly
reduced. Furthermore, within 30 percent RAS binder, not only is the linear blending
chart applicable, but the regular DSR and BBR can also be used to evaluate the high and
low PG temperatures of the blended binders.
Apparently, the work presented in this chapter needs to be further verified and expanded
to include more virgin, RAP, and RAS binders for blending. The blending discussed in this
chapter was done in the lab between virgin binders and extracted/recovered RAS (RAP) binders
so that the blending is 100 percent. However, this may not be the case in the plant as explained
in Chapter 2 (see Figure 9). Therefore, it is critical to evaluate the engineering properties of full
mixes with RAS, which is presented in next chapter.
27
CHAPTER 4
LABORATORY EVALUATION OF RAS MIX ENGINEERING
PROPERTIES
The discussions in last two chapters focused on the extracted and recovered RAS binder
properties and the blending with virgin binders. Binders are just part of the whole mix and the
mix engineering properties rather than binder itself have more impact on field performance.
Thus, this chapter will investigate RAS mix engineering properties. Specifically three subjects
will be discussed: 1) mixing and compaction temperatures for RAS mixes, 2) impact of RAS on
engineering properties of HMA, and 3) actual blending between virgin and RAS binder. Each of
the subjects will be described below.
PRELIMINARY INVESTIGATION OF LAB COMPACTION TEMPERATURE FOR

RAS MIXES
For all virgin mixes, TxDOT has established clearly the mixing and compaction
temperatures corresponding to virgin binders. For example, the mixing and compaction
temperatures for a virgin mix with a PG70-22 binder are 149°C (300°F) and 135°C (275°F).
However, this is not the case for mixes with RAS (or RAP). The mixing temperature and
especially the compaction temperature and curing time have influence on the optimum asphalt
content (OAC), rutting/moisture, and cracking resistance of the mix. Therefore, it is important to
investigate this issue and establish some preliminary guidelines.
Currently, the curing time for all HMA mixes in Texas is 2 hours. This study kept the
same 2 hour curing time as the current practice, but the emphasis was on the compaction
temperature. One of the ways to defining the compaction temperature is to make the lab-mixed
and compacted specimens have similar engineering properties to those of field cores. Based on
this principle, the researchers evaluated a mix with 5 percent RAS used in one of the field test
sections on FM973 in Austin District (more information on FM973 is presented in Chapter 6). It
is a dense-graded Type C mix with a PG64-22 virgin binder and 5 percent TOAS-E. The high
temperature PG of the extracted/recovered binder from the TOAS-E is PG166. The OAC of this
RAS mix is 5.2 percent. Three most often used compaction temperatures in Texas were
evaluated: 121°C (250°F), 135°C (275°F), and 149°C (300°F). Both Hamburg wheel tracking
test and Overlay Test (OT) were performed on both lab-mixed and compacted specimens and
field cores, following TxDOT test procedures (Tex-242-F and Tex-248-F).
Figure 19 shows the test results. It is very clear that the compaction temperature of
121°C (250°F) is too low, since Hamburg rut depth at such temperature is significantly higher
than that of field cores. When comparing the OT cycles, the compaction temperature of 149°C
(300°F) seems too high, because the OT cycles are only a little bit of over half of the field cores.
When considering both Hamburg and OT results, it seems that 135°C (275°F) is a right choice
for compaction temperature of mixes with RAS. These results are preliminary and more
investigation is still needed in this area.
29
14
Hamburg RD@10000 (mm) 12
10
5
OT Cycles
Figure 19. Hamburg and OT Testing Results: Field Core vs. Lab-Mixed and Compacted
Specimens at 3 Compaction Temperatures.
30
IMPACT OF RAS ON MIX PROPERTIES
The inclusion of RAS materials into asphalt mixes often stiffens the mixes with higher
dynamic modulus, and improves the resistance to rutting, but it may greatly jeopardize the
resistance to cracking. Additionally, TOAS and MWAS may have different impact on mix
performance since TOAS has significantly different high temperature grade from MWAS, as
clearly shown in Figure 7. Therefore, it is necessary to thoroughly investigate the impact of RAS
content and types (TOAS/MWAS) on mix engineering properties in terms of dynamic modulus,
rutting, and cracking. In this study, two RAS types (TOAS-E and MWAS-C) and three RAS
percentages (0 percent, 3 percent, and 5 percent) were considered. Note that TOAS-E is very
similar to MWAS-C in terms of RAS aggregate gradation and RAS binder content, and the only
difference between them is RAS binder high PG temperature: TOAS-E=166°C vs. MWAS-
C=122°C. A total of six mixes with the same raw aggregates and similar gradations, as listed in
Table 1, were evaluated under the dynamic modulus test, Hamburg wheel tracking test (HWTT),
and OT. The 0 percent RAS/PG64-22 mix was the control mix. The 0 percent RAS/PG70-22
mix was added to compare with the mixes with 5 percent RAS/PG64-22, since many DOTs
allow virgin binder one grade “dump” when 5 percent RAS is used.
Table 1. Six RAS Mixes Investigated under This Study.

RAS RAS Percentage1 and Virgin Binder
Type
0% RAS/PG70-22 0% RAS/PG64-22 3% RAS/PG64-22 5% RAS/PG64-22
TOAS-E X X
X X
MWAS-C X X
Note: 1- The percentage here is referred as to RAS rather than RAS binder.
The following steps were followed to investigate the impact of RAS contents and type on
OAC and engineering properties of asphalt mixes:
• Step 1: Fix the RAS content (i.e., 5 percent) and adjust virgin aggregates percentage to
make the total aggregates gradation for each RAS mix as close to each other as possible
(Figure 20).
31
Aggregates Gradations of RAS Mixes
100.0
90.0
80.0
70.0
60.0
0%RAS
50.0
3%RAS
40.0
5%RAS
30.0
20.0
10.0
0.0
#200 #50 #30 #8 #4 3/8" 3/4"
Figure 20. Gradations of HMA Mixes with Varying RAS Contents.
• Step 2: Design the RAS mixes and select an OAC following TxDOT’s standard mix
design procedure (Tex-204-F) for dense graded mixes, which are widely used in Texas
(75 percent of all the HMA used in Texas).
• Step 3: Evaluate the dynamic modulus (or stiffness), rutting/moisture resistance, and
cracking resistance of mixes with varying RAS content at its specific OAC.
o Dynamic modulus of each mix was measured following the AASHTO TP79,
“Determining the Dynamic Modulus and Flow Number for Hot Mix Asphalt (HMA)
Using the Asphalt Mixture Performance Tester (AMPT).” The 4 inch (100 mm)
diameter by 6 inch (150 mm) tall specimens with 7 ± 0.5 percent air voids were
fabricated in accordance with AASHTO PP 60, “Preparation of Cylindrical
Performance Test Specimens Using the Superpave Gyratory Compactor (SGC).” For
each mix, two replicates were tested and the average value is reported here.
o Rutting/moisture resistance of RAS mixes was characterized using the Hamburg
Wheel Tracking Test (Tex-242-F). The specimen size of the HWTT is 6 inch
(150 mm) diameter by 2.5 inch (62 mm) height and its target air voids content is
7±1 percent. The HWTT is conducted in a water bath at a constant temperature of
122°F (50°C). The specimens are tested under a rolling 1.85 inch (47 mm) wide steel
wheel using a 158 lb (705 N) force. An average rut depth measured at several
locations including the center of the wheel travel path is reported at end of the test.
o Cracking resistance of RAS mixes was determined using Texas Overlay Tester (Tex-
248-F). The standard specimen size of OT is 6 inch (150 mm) long by 3 inch
(75 mm) wide by 1.5 inch (38 mm) high and its target air voids content is 7±1 percent
after cutting. The OT is run in a displacement controlled mode with a maximum
opening displacement of 0.025 inch (0.63 mm) at test temperature of 77°F (25°C).
The number of cycles to failure (93 percent reduction of the cyclic maximum load
32
from the one measured at the first load cycle) is used as an indicator for cracking
resistance. Note that five replicates of OT specimens were tested for each mix and
the average value of OT cycles is used for comparison. The correlation between OT
result and field cracking performance has been well documented (Zhou and Scullion,
2005; Zhou et al., 2007a), and the OT has been used for evaluating both reflective and
fatigue cracking by different researchers (Bennert et al., 2008; Zhou et al., 2009;
Zhou et al., 2010; Mogawer et al., 2011).
Impact of RAS on OAC of HMA Mixes
Three mix designs were performed for mixes with TOAS-E: 0 percent TOAS-E/PG64-
22, 3 percent TOAS-E/PG64-22, and 5 percent TOAS-E/PG64-22. The OAC for each mix was
determined based on 97 percent density (or 3 percent air voids) and is presented in Table 2. Also
mix designs for mixes with 3 percent MWAS-C/PG64-22, and 5 percent MWAS-C/PG64-22
were conducted as well. It was found that the OACs for the mixes with 3 percent and 5 percent
MWAS-C are very close to those of the mixes with 3 percent and 5 percent TOAS-E. Thus, the
same OAC was selected for the mixes with the same amount of RAS, regardless of TOAS-E or
MWAS-C. Additionally, the OAC of the mix with 0 percent RAS/PG70-22 was kept the same as
that of the mix with 0 percent RAS/PG64-22, since these two mixes have exactly the same raw
aggregates and gradation and the influence of binder type is considered through mixing and
compaction temperatures. In summary, Table 2 lists the OAC of each RAS mixes evaluated
under this section.
It can be seen clearly that with the higher RAS content, the OAC increases. The reason
for this is that the increasing RAS content increases the composite PG grade of the blended
RAS/virgin binder. Therefore, with the higher composite PG grade the mixing and compaction
temperatures should be increased for high RAS mixes. When the mixing and compaction
temperatures are kept the same for each RAS mix, higher RAS mixes will need more asphalt
binder to achieve the same density. The higher OAC somehow offsets the impact of higher RAS
content on engineering properties of RAS mixes, as discussed next.
Table 2. OAC of Each RAS Mix.

RAS Optimum Asphalt Content (%)
Type
0% RAS/PG70-22 0% RAS/PG64-22 3% RAS/PG64-22 5% RAS/PG64-22
TOAS-E
4.7 4.7 4.9 5.2
MWAS-C
Impact of RAS on Dynamic Modulus
Dynamic moduli of each mix measured at different temperatures are shifted following the
time-temperature superposition principle and presented in master curve format. Figure 21 shows
master curves of all six mixes. Overall, the dynamic modulus master curves of these six mixes
33
are close, except that the 0 percent RAS/PG70-22 mix has a little bit higher moduli. Higher RAS
content does not always mean higher stiffness, which depends on the OAC as well.
Additionally, in terms of dynamic moduli of HMA mixes, the TOAS-E and MWAS-C have very
similar impact.
HMA Mixes with TOAS-E

100000
5% RAS/PG64-22 3% RAS/PG64-22
0% RAS/PG64-22 0%RAS/PG70-22
10000
Dynamic Modulus (MPa)
1000
100
10
1E-5 1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3
Reduced Frequency (Hz)
HMA Mixes with MWAS-C

100000
5%RAS/PG64-22 3%RAS/PG64-22
0%RAS/PG64-22 0%RAS/PG70-22
10000
1000
100
10
1E-5 1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3
Figure 21. Impact of RAS on Dynamic Moduli of HMA Mixes.
Impact of RAS on Rutting Resistance
Figure 22 presents the HWTT test results. For those mixes with the PG64-22 binder,
adding RAS improved rutting/moisture damage, regardless of TOAS-E or MWAS-C. However,
34
such improvement is not significant enough to match the impact of the 0 percent RAS/PG70-22
binder. This observation implies that the actual blending between PG64-22 virgin binder and
RAS binder in the mixing and curing process before compaction is different from the pure
binders blending between PG64-22 binder and the extracted/recovered RAS binders (see
Figures 2 and 4). One may take the mix with 5 percent RAS (TOAS-E)/PG64-22 as an example.
If the complete (or 100 percent) blending occurs like those shown Figure 2, the blended binder is
around PG80-15, which is far stiffer than PG70-22 binder, and accordingly the mix with
5 percent TOAS-E should have less rut depth. Surely there is not 100 percent blending in the
mixing, curing, and then compaction processes, and this is not unexpected since the high
temperature grade of the TOAS-E binder is 166°C.
Additionally, Figure 8 shows that the mixes with the TOAS-E have less rut depths than
those with the MWAS-C, which is reasonable since the TOAS-E binder is stiffer than the
MWAS-C binder (see Figure 7).
Impact of RAS Content on Rutting/Moisture Damage

0%RAS/PG70-22 0%RAS/PG64-22
14
3%RAS/PG64-22 5%RAS/PG64-22
HWTT Rut Depth @10000
12
10
Passes (mm)
0
TOAS-E MWAS-C
Figure 22. Impact of RAS on Rutting/Moisture Damage of HMA Mixes.
Impact of RAS on Cracking Resistance
Figure 23 shows the impact of RAS on cracking resistance of HMA mixes. For those
mixes with PG64-22 binder, the use of RAS decreases cracking resistance of HMA mixes,
regardless of TOAS-E or MWAS-C. The TOAS-E has worse effect than the MWAS-C, which
implies that the blending between the virgin binder PG64-22 and the TOAS-E binder (or the
MWAS-C binder) actually occurred. If there is no blending at all, both the mixes with the
TOAS-E should have the same OT cycles as those with the MWAS-C. Although it is unknown
how much the blending actually occurred, one known thing is that the mixes with the MWAS-C
have much better cracking resistance than those with the TOAS-E. This finding is consistent
with the binder blending charts presented previously in which the impact of the MWAS-C on the
low temperature grade of blended binder is much less than that of the TOAS-E (see low
temperature grades of Figures 10 and 12).
35
Compared with the 0 percent RAS/PG70-22 mix, those PG64-22 mixes with either
3 percent or 5 percent RAS have worse cracking resistance. This is not a surprise because if the
blending between virgin PG64-22 binder and RAS binder occurs, the low temperature grade
would become warmer; conversely, if there is no or very limited blending, the total effective
asphalt binder amounts within those RAS mixes would be much less than the 0 percent RAS
mixes with PG70-22 or PG64-22 binder. Regardless, the cracking resistance is a big concern for
those RAS mixes.
Impact of RAS Content on Cracking

0%RAS/PG70-22 0%RAS/PG64-22
50
3%RAS/PG64-22 5%RAS/PG64-22
40
OT Cycles
30
20
10
0
TOAS-E MWAS-C
Figure 23. Impact of RAS on Cracking Resistance of HMA Mixes.
INVESTIGATION OF ACTUAL BLENDING BETWEEN VIRGIN AND RAS BINDER

THROUGH MEASURING ENGINEERING PROPERTIES
The actual blending issue between virgin binder and RAS (or RAP) binder has been
studied and debated for long time. There are three different opinions among different people:
black rock (no blending at all or 0 percent blending), 100 percent blending, and in between. Up
to now, there is no test available to accurately determine the actual blending. In last several
years, there have been at least three interesting studies performed by D' Angelo et al. (2012),
Bonaquist (2007), and Copeland et al. (2010) in this area. All these three studies still could not
tell us how the blending actually occurs, however they provided some insights on this issue.
Therefore, the researchers described these three studies first, and then present the work done in
this area under this project.
Three Recent Studies on RAP Binder Blending in the Literature
• D' Angelo et al. recently conducted a simple blending study between RAP and virgin
binder using size exclusion. Aggregate size exclusion means that the RAP will have a
designated size in the mix and the virgin aggregates will have a different designated size.
D' Angelo et al. employed HMA mixes contained only two distinct fractions (RAP and
virgin aggregates) that can be easily separated by sieving through appropriate sieves after
mixing with virgin binder. The separate of the individual RAP portion from the virgin
36
after blending allows one to investigate whether or not the asphalt binder content is the
same for both portions. If the RAP portion has higher binder content than the virgin
aggregates portion, then the binder in the RAP did not melt and not much blending occurs
between RAP binder and virgin binder. Instead, the virgin binder just coats the RAP
(aggregates and binder) as a whole. The RAP itself before blending had a binder content
of 3.29 percent, and after the blending, the asphalt binder content of the total mix before
separation was 5.0 percent. Figure 24 shows the asphalt binder contents of RAP and
virgin aggregates. RAP has much higher asphalt binder content than the virgin
aggregates. As noted previously, if the RAP binder completely melts and 100 percent
blending with virgin binder occurs, the RAP should have very similar binder content to
the virgin aggregates and the asphalt binder content should be close to 5.0 percent. The
most reasonable explanation is that the binder in the RAP did not melt and the virgin
binder just coats the RAP as a whole. Therefore, it is more reasonable to treat the mixes
with RAP as a composite material (see Figure 9 in Chapter 2).
7
Retained Binder Content (%)
6 5.63
5
4.02
4
3
2
1
0
Separated RAP after Blending Separated Virgin Aggregates
after Blending
Figure 24. Retained Binder Contents for the RAP and Virgin Aggregates after Blending.
• Bonaquist developed an approach for evaluating blending of RAP and virgin binder in
2007. This approach includes five steps: 1) measure mix dynamic modulus, E*;
2) extract and recover binder (assumes 100 percent total blending); 3) measure recovered
binder shear modulus using DSR test; 4) estimate E* based on measured G* using Hirsh
model; and 5) compare the estimated E* to measured E*. Overlapping or similar values
indicate good mixing; otherwise, it is not 100 percent blending. Since 2007, Bonaquist’s
approach has been advocated for determining blending issue. However, there are several
facts that are worth noting about this approach:
1. Bonaquist approach cannot tell exactly how much blending occurs if there is not
100 percent blending.
2. Bonaquist approach requires to extract and recover asphalt binder from the E*
specimen after the E* test. It is well known that the asphalt binder extraction and
recovery can potentially change the binder properties and consequently, it may lead to
37
wrong conclusions on blending determination. Additionally, the extracted/recovered
asphalt binder is 100 percent blending, and it does not represent the actual blending in
the mix.
3. Dynamic modulus (E*) is an important property of asphalt mix. However, it
measures the response of asphalt mix under loading. Even if there is no blending at
all, the measured E* values may be close to those of mixes with partial blending, as
discussed next.
• Copeland et al. (2010) recently reported that Bonaquist approach could not be used to
identify the RAP and virgin binder blending. Copeland et al. conducted a side by side
comparison between the normal mixing and compaction and the complete separation
mixing and compaction. A 28 percent RAP mix from Maryland State Highway Agency
was used in their study. Details are described below:
o Normal mixing and compaction: virgin aggregates, RAP, and virgin binder were
mixed using a normal bucket mixer in a normal way; and compacted E* samples after
a 4-hour short term aging. The asphalt mixture performance test (AMPT) was used to
measure E* at different temperatures.
o Separate mixing and compaction: virgin aggregates and virgin binder were mixed
using the normal way, but the RAP was left out. The RAP was mixed with the virgin
materials (virgin aggregates and binder) just before the compaction. All others were
performed exactly the same as those of normal mixing and compaction. Actually in
this case there is no blending at all, since the RAP was separated out in the mixing
process and the following 4-hour aging process.
o The measured E* results under these two completely different mixing and
compaction conditions are shown in Figure 25. There is no difference in terms of E*
between the normal mixing and compaction process and the complete separation
process. Therefore, the measured E* values cannot be used as an indicators for
binder blending.
38
Figure 25. Measured E* Values of 28 Percent RAP Specimens Produced under Different
Conditions.
In review of these three studies, at least three lessons can be learned: 1) in most cases,
RAP binder does not melt and uniformly blend with virgin binder. Instead, virgin binder coats
RAP as a whole and the coated virgin-RAP becomes a composite; 2) blending may not be
necessary to produce properties similar to blended binder due to composite effect; and
consequently; 3) Bonaquist approach may not be a valid approach for determining blending
between RAP and virgin binder.
Investigation of RAS Binder Blending in Terms of E*
Similar to what Copeland et al. did on RAP binder blending in terms of E*, the
researchers investigated E* difference of a RAS mix produced under normal mixing and
compaction and separated mixing and then mixed just before compaction. The RAS mix
contains 5 percent RAS with a PG64-22 virgin binder, and the total optimum asphalt content is
5.2 percent. Figure 26 shows the E* values at two completely different mixing and compaction
processes measured from AMPT test. The observation made by Copeland et al. on RAP is true
for the RAS mix: blending may not be necessary to produce properties similar to blended binder
due to composite effect.
The researchers further checked the Bonaquist approach for estimating blending.
Figure 27 shows the shear modulus master curve of extracted/recovered asphalt binder (including
both virgin and RAS binders) from the E* specimens. The comparison between the measured E*
from AMPT test and the predicted E* from Hirsh model is presented in Figure 28. It seems that
the predicted E* values are all larger than the measured E*. This indicates that the blending is
very little, because this happens only when the RAP binder does not blend with virgin binder and
the measured E* is dominated by the virgin binder. However, as discussed previously, this may
be due to composite effect. Therefore, it is difficult to make conclusions on this issue. As
39
shown in Chapter 2, the shingles binders, regardless of MWAS or TOAS, are very stiff. So the
question, discussed in the next section, is: does one really want 100 percent blending?
E*-Describe response only

100000
Normal blending Separate blending
10000
1000
100
10
1E-5 1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3
Figure 26. Measured E* Values of 5 Percent RAS Specimens Produced under Different
Conditions.
G* Master Curve
1.E+05
1.E+04
1.E+03
0C
1.E+02
G* (kPa)
20C
1.E+01 40C
60C
1.E+00
80C
1.E-01
100C
1.E-02
1.E-05 1.E-03 1.E-01 1.E+01 1.E+03 1.E+05 1.E+07
Frequency (Hz)
Figure 27. Master Curve of Extracted/Recovered Binder.
40
Bonaquist Approch for Blending
14000
12000
Prediced E* (Mpa)
10000
8000
6000
4000
2000
0
0 2000 4000 6000 8000 10000 12000 14000
Measured E* (Mpa)
Figure 28. Measured vs. Predicted E* Values.
Does 100 Percent Blending Really Matter?
The researchers also compared the true 100 percent blending between RAS binder and
virgin binder and the normal mixing and compaction in terms of Hamburg and OT testing. The
100 percent blending was achieved through extracting and recovering RAS binder in advance
and then manually mixing the RAS binder with virgin binder. Figure 29 illustrates mixing and
molding processes. Figure 30 presents the Hamburg and OT test results. Additionally, the same
mix produced in the plant and compacted in the field was also tested under Hamburg and OT
tests. The results also are shown in Figure 30. As seen in Figure 30, 100 percent blending did
not improve the engineering properties of the RAS mix in terms of Hamburg and OT tests.
41
Virgin RAS Virgin Virgin RAS Virgin
binder aggregate binder aggregate
RAS RAS
Mixing and Molding Hamburg binder aggreg
and OT Specimens
Manually (100
(a) Normal Mixing and Molding %) blending
Mixing and Molding Hamburg and OT

(b) Mixing and Molding with 100% Blending
Figure 29. Illustration of Different Mixing and Molding Processes.
Figure 30. Field Core vs. Lab Normal Blending vs. Lab 100 Percent Blending.
Discussion
The data presented above indicated that no procedure is available to accurately determine
the actual blending of RAP/RAS mixes, regardless of measuring E*, Hamburg rut depth, and OT
cycles. Therefore, the question becomes:
• Is it really important to have a method to determine blending or
• Is it more important to ensure that the required engineering properties of RAS mixes
(such as max. Hamburg rut depth, Min. OT cycles) are met in the mix design process?
For the time being, the researchers believe that it is critical and practical to ensure the
designed RAS mixes have acceptable rutting/moisture resistance and cracking resistance through
Hamburg and OT testing.
42
SUMMARY AND DISCUSSION
Reviewing the results presented above, several findings and a discussion are provided
below:
• Comparing with engineering properties of field cores, the laboratory compaction
temperature and curing time for HMA containing RAS are 135°C (275°F) and 2 hour.
• The use of RAS has no significant influence on dynamic moduli of HMA mixes, but
improves their rutting/moisture damage. Meanwhile, adding RAS generally increases
OAC of HMA mixes and the higher OAC corresponding to higher RAS content.
However, RAS mixes have very poor cracking resistance, compared to the 0 percent RAS
mixes with PG64-22 or PG70-22, even though the RAS mixes have higher OAC.
Therefore, cracking resistance is a big concern for the RAS mixes.
• Impact of the TOAS-E is different from that of the MWAS-C in terms of cracking
resistant. Compared with the TOAS-E mixes, the MWAS-C mixes have much better
cracking resistance. This finding is consistent with the results of binders blending
discussed previously.
• The data presented above indicated that no procedure is available to accurately determine
the actual blending of RAP/RAS mixes, regardless of measuring E*, Hamburg rut depth,
and OT cycles. For the time being, the researchers believe that it is critical and practical
to ensure the designed RAS mixes have acceptable rutting/moisture resistance and
cracking resistance through Hamburg and OT testing.
Although the actual blending extent is still unknown, one known fact is that the use of
RAS decreases cracking resistance of HMA mixes, and some remedies need to be explored, as
discussed in next chapter.
43
CHAPTER 5
APPROACHES FOR IMPROVING CRACKING RESISTANCE OF RAS
MIXES
As discussed in Chapter 4, the use of RAS can improve rutting resistance of HMA mixes,
but it causes poor cracking resistance of the mix and, consequently, the poor durability of HMA
mixes. Therefore, some approaches need to be taken to balance the performance of RAS mixes.
In general, there are at least four approaches:
• Reducing RAS usage (i.e., from 5 percent to 3 percent).
• Rejuvenating RAS binder in the mix design process.
• Using soft virgin binders especially on the low temperature grade (i.e., PGXX-28,
PGXX-34).
• Increasing design density (lowering design air voids) or reducing Ndesign.
Naturally, the first choice is to use less RAS. However, the previous results shown in
Figure 23 indicated that reducing RAS usage from 5 percent to 3 percent does not have
significant improvement on cracking resistance. Further reducing the RAS usage will make no
sense in terms of recycling. The second choice is to rejuvenate RAS binder using some
rejuvenating agents. It sounds like a good idea and potentially improves cracking resistance of
RAS mixes (Tran et al., 2012), but there are lots of practical and technical issues when applied to
normal asphalt plant operations. Apparently, more research is needed in the area of rejuvenating
agents and practical application. Thus, this study focused on the last two approaches: using soft
binder and increasing design density and evaluated the effectiveness of these two approaches in
improving cracking resistance of RAS mixes.
USE OF SOFT BINDERS
The same 5 percent RAS/PG64-22 mixes with the TOAS-E and MWAS-C previously
designed were used as control mixes. Two soft binders selected in this study are PG64-28 and
PG64-34. A total of six mixes (2 RAS and 3 virgin binders) listed in Table 3 were evaluated
under dynamic modulus test (AASHTO TP79), HWTT (Tex-242-F), and OT (Tex-248-F). Note
that the same 5.2 percent OAC was used for all six mixes, since the purpose is to investigate the
influence of soft binders. Figures 31, 32, and 33 show the test results.
Table 3. RAS Mixes with Soft Virgin Binders.

RAS 5%RAS/PG64-22 5%RAS/PG64-28 5%RAS/PG64-34
TOAS-E X x x
MWAS-C X x x
Figure 31 shows that RAS mixes with softer binders have slightly lower moduli, but the
difference among these six mixes is very small in terms of dynamic modulus. Meanwhile,
compared with the 5 percent RAS/PG64-22 mix, the use of softer binders improved
rutting/moisture damage, as indicated in Figure 32. The reason for the improvement is that both
45
PG64-28 and PG64-34 are polymer modified binders. As expected, the mixes with the MWAS-
C have deeper rut depth than those with TOAS-E. Figure 33 clearly indicated that it is very
effective to improve cracking resistance of RAS mixes using soft virgin binders. For the cases
presented here, one grade (−6°C) lower can triple the OT cycles of RAS mixes. Additionally,
the mixes with the MWAS-C always have better cracking life than those with the TOAS-E. In
summary, the use of soft binders has no much impact on dynamic moduli of RAS mixes;
whereas, it can improve both rutting and cracking resistance of RAS mixes, especially on
cracking resistance.
HMA Mixes with 5% TOAS-E

100000
PG64-22 PG64-28 PG64-34
10000
1000
100
10
1E-5 1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3
HMA Mixes with 5% MWAS-C

100000
PG64-22 PG64-28 PG64-34
10000
1000
100
10
1E-5 1E-4 1E-3 1E-2 1E-1 1E+0 1E+1 1E+2 1E+3
Figure 31. Impact of Soft Binders on Dynamic Modulus of 5 Percent RAS Mixes.
46
Impact of Soft Binder on Rutting
12
PG64-22 PG64-28 PG64-34
HWTT Rut Depth @10000 10
Passes (mm)
8
0
TOAS-E MWAS-C
Figure 32. Impact of Soft Binders on Rutting/Moisture Damage of 5 Percent RAS Mixes.
Impact of Soft Binder on Cracking

350
PG64-22 PG64-28 PG64-34
300
250
OT Cycles
200
150
100
50
0
TOAS-E MWAS-C
Figure 33. Impact of Soft Binder on Cracking Resistance of 5 Percent RAS Mixes.
INCREASING DESIGN DENSITY
Another simple way to improve cracking resistance of RAS mixes is to add more virgin
binder into the mixes through increasing design density (or lowering the design air voids) when
selecting OAC. (Alternatively, one can reduce Ndesign.) Currently the design density for
selecting OAC of RAP/RAS mixes is 97 percent. To avoid bleeding problem, the maximum
design density should be less than 98 percent. Again, the same 5 percent RAS/PG64-22 mixes
with the TOAS-E and MWAS-C previously designed were used here. Two design densities:
97 percent and 97.7 percent were used, and accordingly the corresponding OACs are 5.2 percent
and 5.7 percent, respectively. Only the HWTT (Tex-242-F), and OT (Tex-248-F) testing was
47
performed, and the dynamic modulus test was omitted since the previous results did not show
much difference among different RAS mixes. Figure 34 shows the test results.
Figure 34 shows that the higher OAC corresponding to increase design density
significantly improves cracking resistance, which is desirable. Meanwhile the higher OAC
makes the RAS mixes more susceptible to potential rutting/moisture damage. Therefore, one
must exercise cautions when improving cracking resistance of RAS mixes through increasing
design density.
Impact of Increasing Design Density on

Rutting/Moisture Damage
16
HWTT Rut Depth @10000 Passes
5.2% AC with 97% design density

14 5.7% AC with 97.7% design density
12
10
8
(mm)
6
4
2
0
TOAS-E MWAS-C
Impact of Increasing Design Density on Cracking

90
5.2% AC with 97% design density
80
5.7% AC with 97.7% design density
70
60
OT Cycles
50
40
30
20
10
0
TOAS-E MWAS-C
Figure 34. Impact of Increasing Design Density on Rutting/Moisture Damage and Cracking
Resistance of RAS Mixes.
SUMMARY
The OT results presented above clearly indicated that both using soft binder and
increasing design density can improve cracking resistance of RAS mixes. When considering
48
rutting/moisture damage of RAS mixes, using soft binder is superior to increasing design density
(see Figures 32 and 34). In order to validate these laboratory test results and these two
approaches, field test sections were constructed in Texas, which is discussed in next section.
49
CHAPTER 6
PERFORMANCE OF FIELD TEST SECTIONS WITH RAS MIXES
This chapter documents all field test sections constructed with a purpose of recording
field performance of RAS mixes and validating the approaches for improving cracking
performance of RAS mixes: using soft binder and increasing design density. Another purpose of
field test sections is to identify the locations RAP/RAS mixes can be used with low risk. A
variety of field test sections were built under this study around Texas. Detailed information
about these test sections and field performance observed so far are presented below.
RAS TEST SECTIONS ON US87, AMARILLO AND OBSERVED FIELD

PERFORMANCE
Two 3-inch asphalt overlay test sections were constructed side by side on the same lane
and traveling direction on US87, Amarillo, Texas, in late October 2010. The main objective of
these two test sections was to validate the effectiveness of increasing design density on
improving cracking resistance of RAS mixes. The RAS mixes used on the two test sections are
exactly the same (aggregates, gradation, virgin binder, and RAS) except the OAC; the OAC for
control section was 4.6 percent corresponding to 96.5 percent design density and the other one
being 5.0 percent corresponding to 97.3 percent design density. Amarillo’s climate is a
temperate semi-arid climate characterized by numerous freeze-thaw cycles and occasional
blizzards during the winter season. Average daily high temperatures of Amarillo range from
48°F (9°C) in January to 92°F (33°C) in July. The US87 in Amarillo, Texas, has medium traffic
with around 5 million ESALs in 20 years. The existing asphalt pavement had severe transverse
cracking before asphalt overlay. The cold weather and severe existing pavement cracking plus
traffic make these two sections a good case study to rapidly validate the effectiveness of
increasing design density on improving cracking resistance of RAS mixes.
After the completion of the construction of these two RAS test sections, three field
surveys have been conducted on Apr. 5, 2011, Dec. 15, 2011, and May 30, 2012, respectively.
So far, no rutting has been observed, but reflective cracking occurred on both test sections
(Figure 35). The development history of the observed reflective cracking is shown in Figure 36.
Note that prior to placing the overlay the number of pre-existing cracks in each section was
documented and mapped. The reflective cracking rate is therefore defined as the ratio of the
number of reflective cracks to the original number of cracks before the 3-inch overlay.
Apparently, increasing design density significantly improved reflective cracking performance of
the RAS mix on US87, which is clearly shown in Figure 36.
51
Section with 96.5% design density Section with 97.3% design density
Figure 35. Observed Reflective Cracking of RAS Test Sections on US87, Amarillo, Texas.
Reflective Cracking Development of RAS Test Sections

50
Higher OAC-97.3%Density
40
OAC-96.5% Density (control section)
30
20
10
Reflective Cracki ng Rate (%)
0
0 5 10 15 20
Months open to traffic
Figure 36. Reflective Cracking Development History of RAS Test Sections on US87,
Amarillo, Texas.
FIELD TEST SECTION ON SH146, HOUSTON AND OBSERVED PERFORMANCE
In contrast to the extreme cold weather in Amarillo, a field test section was constructed
on SH146 in Houston area where the weather is warm. Furthermore, the test section on SH146,
different from the asphalt overlay on US87, is a new construction and the 15 percent
RAP/5 percent RAS mix was used as the top 2-inch surface layer. The main features of this
section are 1) new construction and 2) warm weather. Since the completion of construction on
Oct. 8, 2010, this test section has been monitored three times. The latest survey was conducted
in May 2012. The test section was in perfect condition: no rutting and cracking, as shown in
Figure 37. The researchers will continue to monitor this test section.
52
Figure 37. Perfect Condition of RAP/RAS Test Section on SH146, Houston.
FIELD TEST SECTIONS ON FM973, AUSTIN, TEXAS, AND ASSOCIATED

PERFORMANCE
A comprehensive series of experimental asphalt overlay test sections were constructed on

FM973 near the Austin Bergstrom International Airport. Compared to the cold weather in
Amarillo, the weather in Austin area is warm. Different from US87, this roadway experiences
very heavy truck traffic as it carries traffic from several aggregate quarries and concrete batch
53
plants. A total of nine test sections were built between Dec. 2011 and Jan. 2012. Part of the
objectives of the test sections on FM973 was to evaluate the effectiveness of using soft binder on
improving cracking resistance of RAP/RAS mixes. Table 4 lists all the mixes used in field test
sections. The main features of these nine sections are:
• HMA vs. RAP/RAS mixes.
• HMA vs. WMA.
• WMA: Foaming vs. Evotherm additive.
• PG64-22 vs. PG58-28.
Therefore, these test sections provided an opportunity for comparing performance of
HMA mixes with WMA mixes side by side.
Prior to the 2-inch asphalt overlay, the overall pavement condition was not bad, and some
areas had low severity level of longitudinal cracking along the wheel passes. The overall
deflection measured using falling weight deflectometer is around 0.28 mm (11 mils). So the
2-inch asphalt overlay is sitting on a solid foundation.
Since the completion of construction, up to now these nine test sections have been
trafficked for six months. The latest survey was conducted in July 2012, and neither rutting nor
cracking was observed on any test section. As one example, Figure 38 shows the conditions of
Sections 3 and 6 in July 2012. Apparently, more time is needed for these test sections to show
difference among these nine test sections in terms of rutting and cracking. TTI researchers will
continue to monitor the performance of these RAP/RAS test sections.
Table 4. Nine Test Sections on FM973, Austin, Texas.

Section Virgin
Type RAP RAS
No. Binder
1 HMA 70-22 0 0
2 HMA 64-22 30 0
3 HMA 64-22 15 3
4 HMA 64-22 0 5
5 HMA 58-28 30 0
6 HMA 58-28 15 3
7 WMA Foaming 70-22 0 0
8 WMA Evotherm 70-22 0 0
9 WMA Evotherm 64-22 15 3
54
Section 3: Section 6:
Figure 38. Pavement Conditions of RAP/RAS Test Sections 3 and 6 on FM973, Austin,
Texas, in July 2012.
RAP/RAS FIELD TEST SECTIONS ON LOOP 820, FORT WORTH
Most recently, four more field test sections were constructed on Loop 820, Fort Worth
side by side. Table 5 presents detailed information on these four test sections. The main features
of these four test sections are 1) RAP/RAS/WMA with Advera additive, 2) soft virgin binder
without changing the OAC, 3) extra virgin binder without changing virgin binder grade, and 4)
pre-blending WMA additive with processed RAS. Additionally, these four test sections are a
2-inch asphalt overlay over cracked continuously reinforced concrete pavement (CRCP), as
shown in Figure 39. Apparently, these test sections provided opportunity to check the impact of
soft binder and extra virgin binder on rutting and cracking performance of RAP/RAS mixes.
The test sections were built on July 19, 2012. So far no distress has been observed yet.
The researchers will continue to monitor these four test sections.
Table 5. Four Field Test Section on Loop 820.

Test section Virgin OAC (%) WMA additive: Advera
binder
Section 0 PG64-22 5.1 Advera as external additive
Section 1 PG64-22 5.1 Advera pre-blended with processed RAS
55
Figure 39. Existing CRCP Condition and Construction of RAP/RAS Test Sections on Loop
820, Fort Worth.
SUMMARY
A variety of RAP/RAS test sections have been constructed with different features around
Texas. The two main objectives of field test sections are to 1) identify where the RAP/RAS
mixes works with low risk and 2) validate the approaches for improving cracking resistance of
RAP/RAS mixes. Based on the performance of test sections observed so far, two preliminary
findings are made:
56
• RAP/RAS mixes can be used as surface layer with strong foundation for new
constructions without much risk, such as SH146.
• Use of extra virgin binder (or increasing design density) can improve cracking resistance
of RAP/RAS mixes as shown on US87.
Apparently, these two findings are very preliminary and continuously monitoring all
these test sections is necessary and critical for making final conclusions.
57
CHAPTER 7
ENVIRONMENTAL AND COST IMPACTS OF RAS ON ASPHALT
MIXTURES
INTRODUCTION
Pavement construction and rehabilitation are among the largest consumers of natural
resources. So, the use of recycled materials in pavements represents an important opportunity to
conserve both materials and energy. However, neither asphalt paving contractors nor any other
industry, for that matter, recycles materials simply to improve the environment. If the particular
recycling process is not profitable, it is not likely to be sustainable. Considering the current and
predicted future price of virgin asphalt cement, the use of RAS in asphalt paving mixtures offers
significant economic as well as environmental benefits. The primary economic driver for asphalt
shingle recycling is the AC cost savings derived by HMA producers (Krivit, 2005; Johnson et al.,
2010). Thus, the main environmental benefits are avoiding deposition of millions of tons of non-
biodegradable material in precious landfills along with conservation of fossil fuels and
aggregates. Recycling of asphalt shingles is a growing industry that could foster business
opportunities, create jobs, and generate revenue.
Many variables come into play when estimating potential cost savings when using RAS
in asphalt mixtures, both HMA and WMA. Such variables include:
• Type of the asphalt mixture produced.
• Price of neat liquid asphalt.
• Amount (%) of RAS used in the paving mixture.
• Type of RAS used (post-consumer or manufacturing waste).
• Cost of aggregates alternative to those contributed by the RAS.
• Landfill tipping fees.
• Capital cost of equipment for grinding/handling RAS.
• Expenses for acquisition, transporting, processing, and handling RAS.
NAHB (1998) reported that roof installation annually generates an estimated 7 to 10
million tons of post-consumer or TOAS and construction debris. US shingle production plants
generate another 750,000 to 1 million tons of MWAS. So, in 2011, 11 million tons of waste
shingles is likely a very conservative estimate. Hansen (2009) pointed out that this represents
more than 2 million tons of asphalt cement potentially available for use in asphalt paving
mixtures, or a replacement of almost 9 percent of the current national need for liquid asphalt in
pavement construction. Assuming 5 percent AC in a paving mixture and a 1-inch thick layer,
2 million tons of HMA is enough to resurface 108,000 lane-miles of pavement—almost halfway
to the moon or more than 4 times around the equator.
Regrettably, Rahim (2010) reported that only 5 percent of shingle waste is recycled in
new construction. The ever increasing amount of waste shingles poses a significant
environmental and economic concern. As the price of crude petroleum, and thus liquid asphalt,
increases, so will the value of RAS. If a suitable means of reusing most these materials can be
found, then their environmental liability could be significantly diminished.
59
According to Rahim (2010), in 2005, AASHTO adopted a standard specification for
asphalt shingle use in HMA. This national specification guides the use of RAS in HMA and
enables HMA producers to design mixtures that satisfy typical specifications of state and local
transportation agencies. The AASHTO specification permits the use of both MWAS and TOAS.
In 2006, AASHTO adopted a recommended practice, PP 53, to supplement the standard
specification and then updated it in 2009. These provide significant aid and encouragement for
utilizing RAS in HMA.
ENVIRONMENTAL ISSUES
Current published information suggests that recycling of asphalt shingles, particularly in

HMA, is technically feasible and is likely to offer economic, environmental, and maybe even
engineering benefits.
Asbestos
Published literature suggests that asbestos is the number one concern related to the use of
RAS (Hansen, 2009). This is primarily an air emission concern related to liberation of asbestos
fibers during grinding and handling ground TOAS and, thus, creating a serious health hazard
(Rahim, 2010). In fact, major impediments to the recycling of TOAS are environmental and
regulatory concerns, predominantly with regard to asbestos (Marks and Gerald, 1997; NAHB,
1998; ARMA, 1998; Zickell, 2003; Lee et al., 2004; Krivit, 2007). The use of asbestos in
residential shingles was discontinued in the late 1970s in the United States. Therefore, asbestos
has not been used in the manufacture of asphalt shingles for about 30 years, and since the typical
life of roofing shingles is from 12 to 25 years, asbestos is seldom encountered in TOAS or in
other roofing products (Townsend et al., 2007; Krivit, 2007).
McMullin (2007) asserted that, in 1963, the content of asbestos in shingles was only
about 0.02 percent by weight. In 1977, the content of asbestos in shingles was only
0.00016 percent by weight. The chances of finding any asbestos in TOAS is a about 0.8 percent
of shingles (roughly, 0.5 percent show a trace, 0.17 percent show 2 percent, 0.11 percent show
5 percent [based on 1770 samples reported by the Chelsea Center at the University of
Massachusetts in 2003]). As a result of these findings, the State of Maine no longer requires
asbestos testing. Grefe (2007) reported that, after hundreds of tests, less than 1 percent yielded
presence of asbestos.
According to Schroer (2009), the National Emissions Standards for Hazardous Air
Pollutants (NESHAP) under the United States Environmental Protection Agency (EPA) has an
exemption based on these facts. In the Appendix of the Code of Federal Regulations, Section 40,
Subpart M, shingles from 4-plex or smaller residential dwellings are exempt from asbestos
testing, in accordance with local regulations. A few tests have indicated a trace of asbestos but
no measurable levels. Experts believe that the asbestos was contained in mastics used for sealing
joints in roofs and/or rolled roofing (Townsend et al., 2007).
Asbestos was once used in asphalt shingles to act as reinforcement (i.e., a fibrous mat) for
the shingle and also acted as a fireproofing/insulating material. Asbestos was also used in certain
other roofing products. Townsend et al. (2007) presented a summary of information published in
the Federal Register along with data they collected by from other sources showing how asbestos
was used in shingles and other roofing products (Table 6).
60
Table 6. Asbestos-Containing Asphalt Roofing Products (after Townsend et al., 2007).
Manufacturer Years Manufactured Product
Barber Asphalt Information not available Asphalt-asbestos roofing
Corporation felt or mat
Carey Manufacturing Information not available Asphalt-asbestos shingles,
Company asbestos finish felt, mastic
The Celotex Corporation 1906 through 1984 Asphalt roof coating and
other miscellaneous
materials
Fibreboard Corporation 1920 to 1968 Roof paint, roll roofing
with asbestos-containing
base sheets, caulking
compounds, plastic
cements, taping and
finishing compounds
General Aniline and Film Information not available Roofing asphalt
Corporation
Johns-Manville 1891 through 1983 Asphalt-asbestos shingles,
Corporation rag-felt shingles, fibrous
roof coating, shingle tab
cement, roof putty
Kaylite Company Information not available Asbestos surface coating
for shingles
National Gypsum 1941 through 1981 Roofing and shingles
Company
Monroe Company Information not available Asbestos surface coatings
for shingles
Rhone-Poulenc Ag Early 1930s through 1976 Adhesives, coatings,
Company sealants, and mastics
United States Gypsum 1930 through 1977 Paper and felt
Company
Most shingle processors in Missouri document the source of the shingles but do not
routinely test for asbestos, following the NESHAP guidance (Schroer, 2009). Hansen (2009)
recaps that the occurrence of asbestos in TOAS from residential roofs will be minimal, but that
the recycling facility operator may expect to occasionally encounter asbestos-containing material
and should be prepared to inspect and manage such materials.
Polycyclic Aromatic Hydrocarbons
Townsend et al. (2007) published an extensive review of environmental issues associated

with use of RAS in HMA. They explained that, since asphalt shingles contain a petroleum-
derived product (i.e., asphalt), they contain polycyclic aromatic hydrocarbons (PAHs). PAHs
comprise a group of more than 100 chemicals formed primarily during the incomplete burning of
coal, oil, or gas (ARMA, 1998). Many PAHs are harmless. However, at elevated levels of
61
exposure, some PAHs are known to have detrimental effects on human health (e.g., cataracts,
kidney and liver damage, cancer).
The potential risk pathways for PAH compounds from RAS are not well understood
(Rahim, 2010). Issues have been raised regarding PAH migration into ground water (e.g.,
leaching from stockpiles), direct exposure to humans via dust during grinding and handling RAS,
and release during handling at HMA facilities. Therefore, do ground recycled asphalt shingles
pose either a direct exposure risk or a leaching risk, or does the use of RAS in HMA production
impact PAH emissions? Leaching tests by Kreich et al. (2002) (leached using the toxicity
characteristic leaching procedure [TCLP]) indicated that four different asphalt roofing materials
yielded results that were below the detection limit (0.1 mg/L) for 29 selected PAHs. Townsend
et al. (2007) reported two other TCLP leaching studies (using materials from Maine and Florida)
with similar results. Wess et al. (2004) assessed the effects of runoff water from asphalt
pavements in California. Samples collected from water-draining road surfaces were analyzed for
PAHs and selected heavy metals (lead, zinc, cadmium). Results indicated that concentrations of
the PAH analytes in all stream and road runoff samples were below the detection limit of
0.5 μg/L.
Townsend et al. (2007) stated that the question of PAH emissions from HMA plants
using RAS has been raised, but no data exist to suggest that such practices would result in PAH
emissions any different from HMA using virgin asphalt. They deduced that environmental risks
associated with PAH migration appear to be small and comparable to that presented by any
material containing asphalt. They further noted that, on a life-cycle basis, overall emissions may
be reduced because of the energy offsets that using recycled asphalt shingles would provide
versus using exclusively virgin asphalt materials.
Greenhouse Gas Emissions
Cochran (2007) conducted a preliminary analysis of reductions in greenhouse gas (GHG)

emissions due to recycling TOAS. Her analyses indicated that the equivalent of about 0.27 to
0.29 lb of CO2 equivalents are reduced for every ton of tear-off asphalt shingles recycled.
Canada produces about 36 million tons of HMA per year. According to Clapham
(2007), if only 5 percent of this total annual production of HMA used recycled shingles, a
reduction of 108,000 tons of CO2 emission could be achieved.
Other Air Emissions
Hughes (1997) pointed out that there is occasionally consternation that asphalt mixtures
containing recycled materials may not be able to be recycled in the future. A particular concern
is whether air emissions from the HMA facility will thereby be increased. However, since the
generic composition of RAS is essentially the same as that of asphalt mixtures, the recyclability
and air emissions of mixtures containing RAS are not concerns. Since the asphalt in RAS is
typically harder than that in HMA or RAP, particularly that in TOAS, one could argue that HMA
or RAP containing RAS will release fewer volatile organic compounds than conventional HMA
or RAP.
Sengoz and Topal (2005) pointed out that shingle recycling may actually reduce
emissions of potentially hazardous components associated with the mining, production, and
transport of virgin materials (asphalt and aggregates) that they replace. Inevitably, regulatory
62
agencies must provide regulations, policies, and permit conditions that (1) afford protection for
human health and the environment, (2) are appropriate for the risk presented, and (3) are not
unnecessarily severe (and thus inhibit recycling).
Energy Savings
According to Krivit (2007), using RAS in HMA plants results in energy savings from
three sources:
• Reduced use of virgin asphalt cement.
• Reduced energy to dry/heat virgin aggregates in the HMA plant.
• Reduced electricity and other fuel to run the overall HMA plant.
Krivit (2007) further stated that, depending on the logistics of the specific shingle
recycling system compared to the traditional HMA plant based solely on virgin materials, there
could be additional energy savings due to reduced transportation (e.g., if shingles are processed
and used near their source of generation).
Cochran (2007), of the US EPA conducted a preliminary analysis of energy savings of
recycling tear-off shingles and found that the equivalent of about 200 kilowatt-hours of
electricity is saved for every ton of tear-off asphalt shingles recycled. She admitted that this
analysis was very preliminary and should be refined.
Life-Cycle Environmental Impacts
Cochran (2006) conducted a comprehensive life-cycle analysis that compared recycling

of asphalt shingles (separated at the job site or separated at a materials recovery facility) with
disposal (in an unlined or lined landfill). This study evaluated environmental impacts from
management methods and emissions to air, soil, and water. Impacts analyzed included global
warming potential, human toxicity potential, abiotic (e.g., water, sand, or gravel) depletion
potential, and acidification potential. According to her analysis, shingle recycling reduced the
environmental and energy burden associated with the manufacture of asphalt from crude oil, but
of course, added some burden as a result of the requirement for processing the shingles prior to
reuse. She found that the net energy requirement associated with recycling shingles into HMA
was less than the requirement associated with disposing of those shingles in a landfill and using
all virgin materials for HMA production.
The University of California at Berkeley developed software (Pavement Life-Cycle
Assessment Tool for Environmental and Economic Effects [PaLATE]) to assist with assessment
of environmental and economic effects of pavements and roads. PaLATE® takes user input for
the design, initial construction, maintenance, equipment, and costs for a roadway and estimates
life-cycle environmental effects and costs. Environmental effects that are investigated include
(http://www.ce.berkeley.edu/~horvath/palate.html):
• Energy consumption.
• CO2 emissions.
• NOx emissions.
• PM10 emissions.
• SO2 emissions.
63
• CO emissions.
• Leachate information.
Waste Reduction Model
The US EPA created a computer program called Waste Reduction Model (WARM) to
help solid waste planners and organizations track and voluntarily report GHG emissions
reductions from several different waste management practices including use of RAS. WARM is
available free both as a web-based calculator and as a Microsoft Excel spreadsheet at:
http://www.epa.gov/climatechange/wycd/waste/calculators/Warm_home.html. The Excel-based
version of WARM offers more functionality than the web-based calculator. WARM calculates
and totals GHG emissions of baseline and alternative waste management practices, e.g., source
reduction, recycling, combustion, composting, and landfilling. The model calculates emissions
in metric tons of carbon equivalent, metric tons of carbon dioxide equivalent, and energy units
(million BTU) across a wide range of material types commonly found in municipal solid waste.
ECONOMIC FACTORS
According to Krivit (2007), the economics of TOAS recycling are currently driven by
three main factors: (1) prevailing landfill tipping fees; (2) price of virgin AC; and (3) cost of
RAS production. The virgin AC price, as a world commodity, will generally follow
national/international trends. The future of AC costs is expected to continue to increase over the
long term. An illustration of the price trend for AC is depicted in Figure 40 (Peterson, 2011).
Note that the prices in Figure 40 are rack prices; to estimate typical bid prices for asphalt, one
should add about $100 or a little more to these values. The point here is that, as the price of
asphalt increases, so does the value of asphalt shingles.
Figure 40. Price Trends for Liquid Asphalt from 1999 to 2011 (after Peterson, 2011).
64
Using conservative values, if one assumes RAS contains 20 percent recoverable AC, the
use of only 2.5 percent RAS in HMA can reduce the virgin binder content by 0.5 percentage
points. That is, one would use 4.5 percent virgin binder instead of 5.0 percent in a typical HMA.
So, for each ton of HMA, 10 lb of virgin liquid AC would be saved. At a production rate of
300 tons per hour for 10 hours, 3000 tons of HMA is produced. Thus, in a day’s production, one
could save 15 tons of virgin AC. Using a cost of $600 per ton for virgin AC, this equates to a
savings of $9000 per day minus the cost of 75 tons of ground shingles. Assuming a total cost of
$40/ton for processing, transporting, and blending the RAS, 2.5 percent ground shingles in HMA
will yield a net savings of $6000 per day or $2.00 per ton. If 5 percent RAS can be
accommodated in the asphalt mixture, obviously, the net savings is double this value.
According to Krivit (2007), Bituminous Roadways, Inc. (BRI), in Missouri, indicated that
the use of MWAS has become their standard practice with a large percentage of the HMA
production incorporating RAS in the same manner as RAP is incorporated (Peterson, 2004). BRI
realized savings of approximately $0.50 to $1.00 per ton of final HMA product with the use of
5 percent RAS. The average cost per ton for HMA was approximately $30 per ton in 2003
(Peterson, 2004). This is very similar to the savings reported by Allied Blacktop, based in Eau
Claire, Wisconsin, who reported savings of about $0.50 per ton of HMA (Ayers, 2003). Maupin
(2008) estimated a savings of $2.69 per ton of HMA containing 5 percent shingles. Other studies
have indicated a savings of up to $3 per ton of final HMA (NAPA, 2000) back in 2000 when AC
was less than $200/ton. With the current cost of AC at about $600 per ton, a savings of $3 per
ton of HMA appears reasonable when using 5 percent shingles, thus, a day’s production of 3000
tons of HMA could save $9000.
Brock (2007), of Astec Industries, used the values for composition of three different
types of shingles, shown in Table 7, and prepared a simplified economic analysis of RAS in
HMA when the cost of liquid AC was $400 per ton (Table 8). Therefore, in 2011, with the cost
of AC now around $600/ton, the cost savings would be significantly greater. Although the value
of RAS has surely increased with the price of AC, the price of RAS has probably not escalated to
the same degree as virgin AC.
Table 7. Typical Shingle Composition (modified after Brock, 2007).

Item New Organic New Fiberglass Old Shingles (TOAS)
Shingles Shingles
lb/100 ft2 % lb/100 ft2 % lb/100 ft2 %
Asphalt 68 30 38 19 73 31
Filler 58 26 83 40 58 25
Granules 75 33 79 38 75 32
Mat -- -- 4 2 -- --
Felt 22 10 -- -- 28 12
Cut-Out (2) 1 (2) 1 0 0
TOTAL 221 -- 202 -- 235 --
65
Table 8. Typical HMA Plant Economic Savings When Using RAS (after Brock, 2007).
It currently appears that the price of liquid AC will increase at a faster rate than that of
RAS. Therefore, the potential savings by using RAS in HMA is expected to increase. Investing
in equipment and training necessary to incorporate RAS into HMA should be prudent. If landfill
operators increased tipping fees for recyclable materials, such as asphalt shingles, economics
would be further pushed in the direction of recycling. Grefe (2007) affirmed that the LaCrosse
County, Wisconsin, landfill uses a differential fee structure to encourage contractors to supply
separated shingles to a processing area.
Krivit (2007) stated that landfill tipping fees vary by region within the United States and
concluded that the economics of shingle recycling are much more favorable on the North East
region of the U.S., where landfill tipping fees can exceed $100 per ton. Most other parts of the
country report tipping fees ranging from less than $10 per ton to about $45 per ton. According to
the TCEQ (2007), the average state-wide tipping fee in Texas in 2006 was $25.70 per ton.
Therefore, Krivit (2007) concluded that shingle disposal is often cheaper than recovery
for several reasons, including:
• Labor costs for sorting.
• Capital costs for processing equipment.
• Relatively low cost of disposal.
• Low market values for recovered products.
• Shingle transportation costs (particularly to rural areas).
Further, the total volumes of recoverable construction and demolition material delivered
to a recycling facility may be relatively low (particularly in Greater Minnesota). Furthermore,
66
these tonnages and economies of scale are even lower when attempting to recover just one
marketable material, such as RAS.
Hughes (1997) originally presented a very simplified economic analysis or worksheet for
recycling RAS into HMA based on a generic cost-benefit model (Table 9). Krivit (2007) added
Item G, in Table 3, to account for capital costs, as he contended that any specific net savings
calculations must include calculations for the budgeted capital (e.g., land, buildings, and
equipment) and operations/maintenance costs. He included additional items in Item H, stating
that cost estimates should include labor, sales, and other marketing costs as well as utilities
(including water) and transportation. Krivit (2007) further commented that all QA/QC costs
related to RAS must be included, along with any laboratory costs for asbestos testing and final
product engineering tests. A notable economic benefit to using MWAS over TOAS is that
asbestos testing is not required.
Hansen (2009) illustrated potential savings of using RAS in an asphalt mixture by using
the following values.
• RAS in mix = 5 percent.
• Effective (recoverable) AC content of RAS = 20%.
• Virgin AC = $600/ton.
• Fine aggregate in shingles = 30%.
• Value of fine aggregate = $10/ton.
• Tipping fee = $25/ton.
• Acquisition cost = $0/ton (assumes generator of waste pays this cost).
• Additional processing cost = $12/ton.
• Capital costs = $0 in this example.
• Miscellaneous costs = $0 in this example.
Using these above values with the original NAPA worksheet (Hughes, 1997), Hansen
(2009) achieved a cost savings of $6.80 per ton of HMA. This clearly illustrates the value of
RAS in asphalt mixtures (both HMA and WMA), with a majority of the savings coming from
replacing virgin AC. The analysis should take into account the higher cost of AC when using
quantities of RAS or RAS/RAP combinations that require a softer than normal grade of AC. In
this case, the savings for Rows A and B would be based on the AC and aggregate content of the
RAS or the combined RAP/RAS and the cost of the standard grade of AC. Then, an additional
cost item would be needed to account for the higher cost of the softer AC. This higher cost
would be the difference between the costs of the softer and standard grade of AC multiplied by
the amount of virgin AC required.
67
Table 9. Method for Calculating the Value of RAS in Asphalt Mixtures
(modified after Hughes, 1997; Hansen, 2009; Krivit, 2007).
________________________________________________________________________
Calculation $/ton of
Finished HMA
Savings
A. Savings from reduced need for new (virgin) asphalt cement (AC)
New AC $/ton ( ) x %AC in RAS ( ) x % RAS in mix ( ) $__________
B. Savings from new (virgin) fine, bituminous aggregate
New fine agg. $/ton ( ) x % fine agg. in RAS ( ) x % RAS in mix ( ) $__________
C. Savings from tipping fee
Tipping fee $/ton ( ) x % RAS in mix ( ) $__________
D. Total Gross Savings per ton of hot mix (add: A + B + C) = $__________
Costs
E. Less acquisition cost of RAS (e.g., trucking cost):
Acquisition cost $/ton ( ) x % of RAS in mix ( ) $__________
F. Less additional processing costs (e.g., sorting, crushing, screening):
Processing cost $/ton ( ) x % of RAS in mix ( ) $__________
G. Less capital costs (e.g., equipment, land, improvements)
Capital costs $/ton ( ) x % of RAS in mix ( ) $__________
H. Other miscellaneous costs of testing, engineering design
(e.g., asbestos monitoring, mix design, other QC/QA)
Costs $/ton ( ) x % RAS in mix ( ) $__________
I. Total costs (add: E + F + G + H) = $__________
Net savings per ton of hot mix asphalt (Subtract: D − I) = $__________

________________________________________________________________________
68
Rand (2011) deduced that proper use of unmodified binders (e.g., PG 64-22 instead of
PG 70-22 or PG 76-22) along with RAP and RAS can reduce the cost of asphalt pavement
material by more than $15/ton. The assumptions shown in Table 10 were used to determine the
HMA cost estimates in Table 11 and Figure 41. Note that the cost estimates in Table 11 and
Figure 41 represent material costs only. These do not reflect the total as-constructed cost of
HMA and are based on 2011 cost data in Texas. Costs can vary significantly with
circumstances.
Table 10. Assumptions Used for Asphalt Pavement Cost Estimates.

Cost Per
Material Notes
Ton
Aggregate $22 Includes processing & freight
PG 76-22 $538 Based on September 2009 *Index (freight not

included)

included)

included)
RAP $15 Contains 5% AC, includes processing & freight
RAS $20 Contains 20% AC, includes processing & freight
*Source: Louisiana Asphalt Pavement Association
Table 11. Asphalt Pavement Cost Estimates.

Binder Virgin 20% RAP 5% RAS 15% RAP + *One Grade
Mix 5% RAS Softer Binder
Grade Only Only
PG 76-22 47.80 41.24 42.54 37.64 35.74
PG 70-22 44.90 38.92 40.22 35.74 32.39
PG 64-22 39.75 34.80 36.10 32.39 NA
*Includes 15% RAP and 5% RAS
69
HMA Cost, $/ton
60
50
40
30
20
10
0
Virgin Mix 20% RAP 5% RAS 15% RAP + 1-Grade 2-Grade
5%RAS Softer Softer
Figure 41. HMA Cost in Dollars per Ton for Type D (1/2-inch NMAS)
Mix Containing PG 76-22.
TxDOT typically uses 5 million to 15 million tons of asphalt mix annually. Assuming
TxDOT could save $10/ton by using RAP, RAS, and unmodified binders, an annual savings of
$50 to $150 million is achievable (Rand, 2011). Robinette and Epps (2010) agree that the value
of RAP and RAS depends on local market conditions, e.g., price of virgin asphalt binder, crushed
virgin aggregate, and processing RAP and/or RAS. Because the price of asphalt binder
fluctuates, the value of RAP and RAS changes almost daily.
PERTINENT ACTIVITIES TO ENCOURAGE RECYCLING
Various government agencies are creating regulations to encourage recycling of shingles.

These activities usually come under the general heading of construction and demolition (C&D)
debris. For example, in 1990, the Massachusetts Department of Environmental Protection
(MassDEP) introduced its first bans on landfilling and combustion of easy-to-recycle and toxic
materials. Additional waste bans have been phased in over time. A few of the materials
prohibited from disposal in Massachusetts that have been used in pavements are:
• Asphalt shingles.
• Asphalt pavement, brick, and concrete.
• Glass and metal containers.
• Whole tires (banned from landfills only; shredded tires acceptable).
Since the first waste bans were introduced, Massachusetts municipalities and businesses,
often supported by MassDEP grants and technical assistance, have developed new infrastructure
to collect banned items and other discarded materials, and to divert them from disposal to reuse
and/or recycling (http://www.mass.gov/dep/recycle/solid/wastebans.htm).
In 2006, the US Army published a memorandum titled Requirements for Sustainable
Management of Waste in Military Construction, Renovation and Demolition Activities. Briefly
stated, this policy mandates that all new construction, renovation, and demolition projects
include contract performance requirements to divert, as a minimum, 50 percent of non-hazardous
C&D debris from landfill disposal. The Army’s goal for C&D debris diversion is based partly
70
on levels considered achievable by other public agencies responsible for solid waste
management. California, City of Chicago, and Nova Scotia (Kenney, 2007) require diversion of
at least 50 percent of C&D waste from construction, remodeling, reproofing, and demolition
projects. City of Halifax requires 75 percent diversion of C&D (Kenney, 2007). Many other
jurisdictions have enacted ordinances to require C&D waste diversion or exclude C&D materials
from landfill disposal. The Federal Leadership in High Performance and Sustainable Buildings
Memorandum of Understanding and Guiding Principles require that at least 50 percent of C&D
debris be recycled or salvaged. So, while the Army’s C&D waste management policy is
progressive, it is not without precedent in the public sector.
http://www.erdc.usace.army.mil/pls/erdcpub/www_org_info.show_page?f_id=2364657&f_parent=55174
SUMMARY AND CONCLUSIONS
These conclusions are based on findings from this review of published information.
• Discarded asphalt shingles may consist of MWAS or TOAS. Recycling of MWAS has
been widely accepted because of its relatively homogeneous nature, lower oxidation, and
freedom from asbestos. TOAS recycling is technically feasible, but the practice has been
limited in some areas because of concerns about asbestos and, to a lesser extent, PAHs.
• Using 5 percent RAS in HMA and assuming a cost of $600 per ton for virgin AC, a
contractor can save about $4.00 to $7.00 per ton of HMA, depending on the cost to
acquire the RAS.
• The net energy requirement associated with recycling shingles into HMA is less than the
requirement associated with disposing of those shingles in a landfill and using all virgin
materials for HMA production.
• Use of RAS in HMA should provide environmental benefits by offsetting the use of
virgin asphalt and by reducing the volume of debris in landfills.
• The occurrence of asbestos in TOAS from residential reroofing projects will be very
limited (and will decrease with time), but the recycling facility operator should expect to
encounter asbestos on occasion and, thus, should be adequately prepared to monitor and
manage such material.
• Risks associated with polycyclic aromatic hydrocarbons (PAH) migration appear to be
small and comparable to that encountered when handling any asphalt-containing material.
Data do not exist to suggest that incorporating RAS into HMA should be limited because
of PAH concerns.
• Using RAS in HMA to conserve virgin asphalt instead of disposing of shingles in
landfills will reduce greenhouse gas emissions.
• Models are available for estimating energy savings, greenhouse gas emissions as well as
life-cycle costs and environmental effects when using RAS in HMA. Some of these
models are identified in this document.
• Future recyclability and air emissions of pavements containing RAS are not concerns.
Asphalt in RAS is typically harder (less volatile) than that in HMA or RAP, particularly
that in TOAS, therefore, HMA or RAP containing RAS should liberate fewer volatile
organic compounds than conventional HMA or RAP.
• Selected public agencies now mandate that all new construction, renovation, and
demolition projects include contract performance requirements to divert some minimum
percentage of non-hazardous C&D debris from landfill disposal.
71
CHAPTER 8
SUMMARY AND CONCLUSIONS
This report presents a comprehensive study on HMA mixes containing RAS, including
RAS binder characterization and blending charts for virgin/RAS binders, impact of RAS content
on OAC and engineering properties of RAS mixes, and approaches for improving cracking
resistance of RAS mixes. Furthermore, a variety of RAS field test sections were constructed to
validate the approaches for improving cracking resistance of RAS mixes. Additionally, this
report discusses the environmental and economic benefit of using RAS in HMA. Based on the
research presented in this report, the following conclusions are offered:
• RAS binders are very stiff. TOAS binders with an average of high temperature grade of
175°C are much stiffer than MWAS binders, which have an average of high temperature
grade of 131°C. The MWAS has smaller variation in the high temperatures grade,
compared to the TOAS varying from 159°C to 214°C.
• Generally, the virgin and RAS binders blending is non-linear. For practical application,
the linear blending chart can still be used if the RAS binder percentage is not beyond
30 percent. Within 30 percent RAS binder, not only is the linear blending chart
applicable, but the regular DSR and BBR can also be used to evaluate the high and low
ends of the PG grade of the blended binder.
• Impact of MWAS binder on the high and low PG temperatures of virgin binder is
different from that of TOAS binders. Compared with the TOAS binders, the MWAS
binders have less impact on PG temperatures of virgin binders, which is reasonable since
MWAS binders are much softer than those TOAS binders. Thus it is important to
consider differentiating MWAS from TOAS when used in asphalt mixes.
• The use of RAS has no significant influence on dynamic moduli of HMA mixes, but
improves their rutting/moisture damage. Meanwhile, adding RAS generally increases
OAC of HMA mixes and the higher OAC corresponding to higher RAS content.
However, RAS mixes have very poor cracking resistance, compared with the 0 percent
RAS mixes with PG64-22 or PG70-22, even though the RAS mixes have higher OAC.
Therefore, cracking resistance is a big concern for the RAS mixes.
• Two approaches for improving cracking resistance of RAS mixes were explored in the
laboratory. The test results clearly indicated that both using soft binder and increasing
design density can improve cracking resistance of RAS mixes. When considering
rutting/moisture damage of RAS mixes, using soft binder is superior to increasing design
density.
• Performance of field test sections observed so far indicated that RAS mixes can be
successfully used as surface layer of new construction with strong foundation. It also has
been observed that the use of extra virgin asphalt binder (or increasing design density)
can effectively improve cracking performance of RAS mixes.
The findings from this study still need to be further validated with field performance data
from different test sections under various scenarios. More work is also needed to investigate the
performance of RAS/RAP mixes produced at the warm mix temperatures.
73
REFERENCES
ARMA, “Polyaromatic Hydrocarbon Emissions from Asphalt Processing and Roofing
Manufacturing Operations,” Asphalt Roofing Manufacturers Association, Rockville, Maryland,
1998.
Ayers, B., “Shingle Recycling: A Paving Contractor’s Perspective,” Second Recycled Asphalt
Shingles Forum, Minneapolis, Minnesota, 2003.
Bennert, T. and A. Maher, Field and Laboratory Evaluation of a Reflective Crack Interlayer in
New Jersey, Transportation Research Record, No. 2084, 2008, pp 112-123.
Bonaquist, R. Can I Run More RAP? HMAT, National Asphalt Pavement Association, Vol. 12,
No. 5, 2007.
Brock, J. D., “From Roofing Shingles to Roads,” Technical Paper T-120, Astec Industries, Inc.,
Chattanooga, Tennessee, 2007.
Button, J.W., D. Williams, and J. Scherocman, “Roofing Shingles and Toner in Asphalt
Mixtures,” Research Report 1344-2F, Texas Transportation Institute, Texas A&M University,
College Station, Texas, 1996.
Clapham, M., “Recycling Roofing Materials in Canada,” presentation at 3rd Asphalt Shingles
Recycling Forum, Minerals and Metals Sector, Natural Resource Canada, Chicago, Illinois, 2007
(accessed 9/20/11 at www.p2pays.org/ref/42/41569.ppt).
Cochran, K., “Construction and Demolition Debris Recycling: Methods, Markets, and Policy,”
Ph.D. Dissertation, University of Florida, Gainesville, 2006.
Cochran, K., “Asphalt Shingles Recycling: A National Perspective,” presented at 3rd Asphalt
Shingles Recycling Forum, U.S. Environmental Protection Agency, Office of Solid Waste,
Chicago, Illinois, 2007.
Copland, A., R. Dongre, and J. D' Angelo, Evaluation of RAP Binder Blending II, presented at
RAP ETG, Oklahoma City, OK, Oct. 27, 2010.
D' Angleo, R. Dongre, and Copland, A., Evaluation of the Extent of Asphalt Blending in RAP
Mixes, Proceedings of the Fifty-Sixth Annual Conference of the Canadian Technical Asphalt
Association (CTAA) - Quebec City, Quebec, 2011, pp.289-308.
Dykes, J., “Recycled Asphalt Shingle Use in Hot Mix Asphalt,” presentation at seminar program
panel session “TH-26—Recycled Asphalt Shingle Use in Hot Mix Asphalt,” ConExpo-Con/Agg
Show and Exposition, Las Vegas, Nevada, 2002.
Epps, J.A. and G.L. Paulsen, “Use of Roofing Wastes in Asphalt Paving Mixtures—Economic
Considerations,” Report 709-2, Center for Construction Materials Research, University of
Nevada-Reno, 1986.
Foo, K. Y., D. I. Hanson, and T. A. Lynn, Evaluation of Roofing Shingles in Hot Mix Asphalt,
ASCE Journal of Materials in Civil Engineering, Vol. 11, No. 1, 1999.
Grefe, R.P., “Wisconsin Regulatory Experience with Asphalt Shingle Recycling,” representing
Bureau of Waste & Materials Management, Wisconsin Department of Natural Resources,
Chicago, Illinois, 2007.
75
Grzybowski, K.F., “Recycled Asphalt Roofing Materials—A Multi-Functional, Low-Cost Hot
Mix Asphalt Pavement Additive,” Use of Waste Materials in HMA Pavement, ASTM Special
Publication 1193, American Society for Testing and Materials, Philadelphia, Pennsylvania,
1993.
Hansen, K., “Guidelines for the Use of Reclaimed Asphalt Shingles in Asphalt Pavements,”
Information Series 136, National Asphalt Pavement Association, Lanham, Maryland, 2009.
Hughes, C.S., Special Report 179, “Uses of Waste Asphalt Shingles in HMA- State of the
Practice,” National Asphalt Pavement Association, Lanham, Maryland, 1997.
Janisch, D.W. and C.M. Turgeon, “Minnesota’s Experience with Scrap Shingles in Bituminous
Pavements,” Report MN/PR-96/34, Office of Materials and Road Research, Minnesota
Department of Transportation, Maplewood, Minnesota, 1996.
Johnson, E., G. Johnson, S. Dai, D. Linell, J. McGraw, and M. Waton, Incorporation of Recycled
Asphalt Shingles in Hot Mixed Asphalt Pavement Mixtures, Minnesota Department of
Transportation, Maplewood, MN, February 2010.
Kenney, B., “Nova Scotia – Too Good To Waste,” representing Nova Scotia Department of
Environmental & Labour, Chicago, Illinois, 2007.
Kriech, A.J., J.T. Kurek, L.V. Osborn, H.L. Wissel, and B.J. Sweeney, “Determination of
Polycyclic Aromatic Compounds in Asphalt and in Corresponding Leachate Water,” Polycyclic
Aromatic Compounds, 22 (3-4), 2002, pp. 517-535.
Krivit, D., “Overcoming the Barriers to Asphalt Shingle Recycling: Final Report for the RMRC
Project 22, Final Report, prepared by Dan Krivit and Associates for the Recycled Materials
Resource Center, University of New Hampshire, Durham, NH, February 2005 (accessed 8/24/11
at http://your.kingcounty.gov/solidwaste/linkup/documents/Overcoming_barriers-shingle.pdf).
Krivit, D., “Recycling Tear-Off Asphalt Shingles: Best Practices Guide,” The Construction
Materials Recycling Association (CMRA), Eola, Illinois, 2007
(http://www.shinglerecycling.org/sites/www.shinglerecycling.org/files/shingle_PDF/ShingleRec
ycling-BPG-DFK-3-22-2010.pdf).
Lee, W. J., W.H. Chao, M.L. Shih, C.H. Tsai, T.J.H. Chen, and P.J. Tsai, “Emissions of
Polycyclic Aromatic Hydrocarbons from Batch Hot Mix Asphalt Plants,” Environmental Science
& Technology, 38(20), 2004, pg. 5274-5280.
Lindsey, R., Owner, Lindsey Contractors, Inc., Waco, Texas, personal communication, August
2011.
Lum, P., “Roofing Shingles in HMA,” 2006 National Conference, Materials and Energy
Conservation for HMA, Indianapolis, Indiana, 2006.
Marks, V. J., and P. Gerald, “Let Me Shingle Your Roadway,” Research Project HR-2079, Iowa
Department of Transportation, Ames, Iowa, 1997.
Maupin, G.W., Jr., “Use of Manufactured Waste Shingles in a Hot Mix Asphalt Field Project,”
Report No. VTRC 08-R11, Virginia Transportation Research Council, Virginia Department of
Transportation, Charlottesville, Virginia, 2008.
76
McGraw, J., A. Zofka, D. Krivit, J. Schroer, R. Olson, and M. Marasteanu, Recycled Asphalt
Shingles in Hot Mix Asphalt, Journal of Association of Asphalt Paving Technologists, Vol. 76,
San Antonio, TX, March 2007.
McMullin, R., ”Third Asphalt Shingle Recycling Forum,” representing Maine Department of
Environmental Protection, Chicago, Illinois, 2007.
NAHB Research Center, “From Roofs to Roads – Recycling Asphalt Roofing Shingles into
Paving Materials,” 1998 (accessed 8/11/11 at
http://www.epa.gov/osw/conserve/rrr/imr/cdm/pubs/roof_br.pdf).
NAPA, Special Report 187, “Recycling Practices for HMA,” National Asphalt Pavement
Association, Lanham, Maryland, 2000.
Newcomb, D., M. Stroup-Gardiner, B. Weikle and A. Drescher, “Properties of Dense-Graded
and Stone Mastic Asphalt Mixtures Containing Roofing Shingles,” Use of Waste Materials in
Hot Mix Asphalt, ASTM Special Technical Publication 1193, ASTM, Conshohocken,
Pennsylvania, 1993.
Paulsen, G., M. Stroup-Gardiner, and J.A. Epps, “Roofing Waste in Asphalt Paving Mixtures –
Economic Considerations,” Report No. 6-331-709-1, Center for Construction Materials
Research, University of Nevada-Reno, 1986.
Peterson, G.D., Manager, Asphalt and Chemical Branch, Texas Department of Transportation,
Austin, Texas, personal communication, August 2011.
Peterson, K., “A Project to Improve a System for Processing Shingle Scrap and to Increase the
Market for Processed Shingles,” Bituminous Roadways, Inc., submitted to Minnesota Office of
Environmental Assistance in completion of a Targeted Market Development Grant for Fiscal
Years 2001-2002 (extended to April 2004), 2004.
Rahim, A., “Evaluation of Waste Recycled Materials Applications in Highway pavements in
California: An Overview of Research,” Department of Civil and Environmental Engineering,
California Polytechnic State University, San Luis Obispo, California, 2010.
Rand, D.E., “High Priced Asphalt inTexas – Going the Way of The Cave Man?,” Hot Mix
Asphalt Technology, Volume 16, No. 2, National Asphalt Pavement Association, Lanham,
Maryland, March/April 2011, pp. 24-26.
Robinette, C. and J. Epps, “Energy, Emissions, Material Conservation, and Prices Associated
with Construction, Rehabilitation, and Material Alternatives for Flexible Pavement,” Journal of
the Transportation Research Board, No. 2179, Transportation Research Board of the National
Academies, Washington, D.C., 2010, pp. 10–22.
Schroer, J., “Asphalt Shingles in HMA: Missouri DOT Experience,” North Central Users and
Producers Group (NCAUPG) Conference, Minneapolis, Minnesota, 2007.
Sengoz, B. and A. Topal, “Use of Asphalt Roofing Shingle Waste in HMA,” Construction and
Building Materials, Vol. 19, No. 5, 2005, pp. 337-346.
Shepherd, P.B., T.J. Powers, J. Hardy, R. Maloof, and C. Patenaude, “Energy Conservation
Through Recycling of Factory Asphalt Roofing Waste,” prepared for the U.S. Department of
Energy, Washington, D.C., 1989.
77
TCEQ, “Municipal Solid Waste in Texas: A Year in Review - 2006 Data Summary and
Analysis,” Texas Commission on Environmental Quality, Waste Permits Division, Austin,
Texas, 2007.
Townsend, T., J. Powell, and C. Xu, “Environmental issues Associated with Asphalt Shingle
Recycling,” Construction Materials Recycling Association, Eola, Illinois, 2007.
Tran, N., A. Taylor, and J. R. Willis, Effect of Rejuvenator on Performance Properties of HMA
Mixtures with High RAP and RAS Contents, NCAT Report 12-05, National Center for Asphalt
Technology, Auburn, AL, 2012.
Wess, J. A., L.D. Olsen, and M.H. Sweeney, “Asphalt (Bitumen),” Concise International
Chemical Assessment Document 59, World Health Organization, Geneva, Switzerland, 2004.
Williams R. C., Cascione A., Haugen D. S., Buttlar W. G., Bentsen R. A., and Behake J.,
Characterization of Hot Mix Asphalt Containing Post-Consumer Recycled Asphalt Shingles and
Fractionated Reclaimed Asphalt Pavement, Iowa State University, Department of Civil,
Construction and Environmental Engineering, Ames, IA, January 2011.
Mogawer, W., A. Austerman, M. E. Kutay, F. Zhou, Evaluation of Binder Elastic Recovery on
HMA Fatigue Cracking using Continuum Damage and Overlay Test Based Analyses,
International Journal of Road Materials and Pavement Design, Vol. 12, No. 2, 2011, pp345-376.
Zhou F. and Scullion T., Overlay tester: a simple performance test for thermal reflective
cracking, Journal of the Association of Asphalt Paving Technologists, Vol. 74, 2005, pp 443-
484.
Zhou F., Hu S., Scullion T., and Chen D., The Overlay Tester: A Simple Performance Test for
Fatigue Cracking, Transportation Research Record, No. 2001, 2007a, pp 1-8.
Zhou F., Hu S., Scullion T., Mikhail M., and Walubita L. F., A Balanced HMA Mix Design
Procedure for Overlays, Journal of the Association of Asphalt Paving Technologists, Vol. 76,
2007b, pp. 823-850.
Zhou F., Hu S., Hu X., and Scullion T., Mechanistic-Empirical Asphalt Overlay Thickness
Design and Analysis System, Research Report FHWA/TX-09/0-5123-3, Texas Transportation
Institute, College Station, TX, October 2009.
Zhou F., Fernando E.G., and Scullion T., Development, Calibration, and Validation of
Performance Prediction Models for the Texas M-E Flexible Pavement Design System, Research
Report FHWA/TX-10/0-5798-2, Texas Transportation Institute, College Station, TX, August
2010.
Zickell, A.J., “Asbestos Analysis of Post-Consumer Asphalt Shingles,” Technical Report 41,
Chelsea Center for Recycling and Economic Development, University of Massachusetts Center
for Environmentally Appropriate Materials, Chelsea, Massachusetts, 2003.
78

0 6614 2 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

0 6614 2 PDF

Uploaded by

Copyright:

Available Formats

Technical Report Documentation Page

1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

7. Author(s) 8. Performing Organization Report No.

Fujie Zhou, Ph.D., P.E.

Sheng Hu, Ph.D., P.E.

Joe W. Button, P.E.

Jon A. Epps, P.E.

Performed in cooperation with the

TEXAS A&M TRANSPORTATION INSTITUTE

Figure 1. Flow Chart of Validation of Binder Extraction and Recovery Methods............................ 4

Table 1. Six RAS Mixes Investigated under This Study. ................................................................ 31

VALIDATION OF ASPHALT BINDER EXTRACTION AND RECOVERY METHODS

MWAS produced with the Solvent:

Binder extraction and recovery:

DSR/BBR Extracted/recovered binders FTIR

Figure 1. Flow Chart of Validation of Binder Extraction and Recovery Methods.

Figure 3. FTIR Test: Trichloroethylene.

DSR: Original Shingles Binder

BBR@0°C: PAV Aged Shingles Binder

BBR@0°C: PAV Aged Shingles Binder

Figure 4. Rheological Properties of Shingles Binder: before and after Extraction/Recovery.

As discussed previously, roofing shingle binders, regardless of MWAS or TOAS, are

Selection of RAS Samples

RAS Binder Extraction and Recovery

RAS Binder Characterization

Figure 5. BBR Test: Early Fractured RAS Binder Beam.

• DSR test results

Figure 7. High Temperature Grades of RAS Binders: MWAS and TOAS.

DISCUSSION ON PRODUCTION TEMPERATURES OF RAS MIXES

Virgin Mix RAS Mix

Figure 9. Illustration of Coating during Production: Virgin vs. RAS/RAP Mixes.

SUMMARY AND FINDINGS

INVESTIGATION OF VIRGIN AND RAS BINDERS BLENDING

Virgin and RAS Binders

Laboratory Testing, Results, and Analysis

Binder Blending: PG64-22-A/TOAS-E Binder

Binder Blending: PG64-22-A/TOAS-E Binder

Figure 10. Binder Blending between PG64-22-A and TOAS-E Binder.

High Temp. Grade (°C) 140

Binder Blending: PG64-28 and TOAS-A Binder

Figure 11. Binder Blending between PG64-28 and TOAS-A Binder.

High Temp. Grade (°C) 100

Binder Blending: PG64-22-B/MWAS-C Binder

Figure 12. Binder Blending between PG64-22-B and MWAS-C Binder.

High Temp. Grade (°C) 120

Binder Blending: PG64-22-B and MWAS-A Binder

Figure 13. Binder Blending between PG64-22-B and MWAS-A Binder.

EVALUATION OF BLENDING AMONG VIRGIN, RAP AND RAS BINDERS

Binder Blending with Fixing 20% TOAS-E Binder:

Binder Blending with Fixing 20% RAP-A Binder:

Binder Blending with Fixing 5% MWAS-A Binder:

100 50%PG64-22+10%RAP +40%RAS 30%PG64-22+10%RAP+60%RAS

Binder Blending with Fixing 10% RAP-B Binder:

DISCUSSION ON MAXIMUM ALLOWABLE ASPHALT BINDER REPLACEMENT

Low Temp. Grade (°C)

PRELIMINARY INVESTIGATION OF LAB COMPACTION TEMPERATURE FOR

Hamburg RD@10000 (mm) 12

Table 1. Six RAS Mixes Investigated under This Study.

Figure 20. Gradations of HMA Mixes with Varying RAS Contents.

Impact of RAS on OAC of HMA Mixes

Table 2. OAC of Each RAS Mix.

Impact of RAS on Dynamic Modulus

HMA Mixes with TOAS-E